Victron BMV Battery Monitor - Peukert Exponent

[wildebus]

The Peukert Exponent? What the hell is that?
It is meant to be a way for a Battery Monitor to adjust the 'real' State of Charge of a battery, adjusting the battery capacity for the load that is being taking out at any given time (more about Peukert here - https://en.wikipedia.org/wiki/Peukert's_law)
If you look at Battery Specs on a site such as Alpha Batteries, you will see different capacities at different discharge rates e.g. the Ritar Lead Carbon batteries I have are rated at 100Ah at C20 (so a current of 1/20th of 100Ah (= 5A) being drawn), but at a rate of C100 (1/100th of 100Ah = 1A), the capacity is 113Ah (13% greater).
So the easier the draw on the battery, the more actual capacity you have in total available.

So what's that got to do with the BMV? Well, the BMV allows you to put in a value for Peukert so the Monitor correctly adjusts for this as the battery is being used. The Monitor shows how many Ahs have been drawn from the battery via the shunt, but to translate that to an accurate State of Charge calculation, the right Peukert value should be entered into the setting.
The right value to use varies by battery type and even to a degree by Battery itself. The higher the value, the more impact it has on the SoC calculation as current draw varies over the day.

On my own Motorhome, I found that the correlation between the BMVs SoC display and the voltage didn't look quite right. At a voltage of 12.02V, I was showing an SoC of 54%. Typically on Lead Acid, 50% SoC on a battery would give an resting voltage of 12.05V, so 54% would be higher
As I have a small load (1.5A), I need to take that into account and the voltage would be a little higher but not more than 12.05V I think. So .... this means the Peukert value I have set is likely a little too high. (it is something that needs tuning in a setup).

Couple of Screenshots to show the difference when changing the number
This is the setting I originally had, with an Exponent of 1.10

Dropping the value a few points, you instantly see the revised SoC value as the calculations are updated.

This initial tuning is needed to make sure you have the right values applied and so getting accurate info from your Victron Monitor.

 

[HarryW]

Slightly confused by the results.
My understanding was the nearer to unity (1) the Pk value was the greater the efficiency hence battery longer battery life?

I realise it’s more complicated than that, just the results seem to be the opposite to what I would have expected?

 

[wildebus]

Slightly confused by the results.
My understanding was the nearer to unity (1) the Pk value was the greater the efficiency hence battery longer battery life?

I realise it’s more complicated than that, just the results seem to be the opposite to what I would have expected?

Not sure what you are meaning by efficiency, or battery life with respect to the Peukert exponent value.

What this result indicates is that as you change the Value, the SOC calculation alters, as it recalculates how the draw current level affects the capacity.
Now if is hard to actually say the results are the way you would expect or opposite to them. Reason being you don't know the type of current this system has been drawing
If I were drawing a tiny current all the time, then the capacity would be greater then the "headline" one the battery maker quotes. if I were drawing a big current all the time, then the capacity would be lower than quoted.
The larger the value over 1, the greater the capacity is impacted by current draw level (and a value of 1.00 means there is no effect).

So to explain further ...
I think the following is why when I lowered the value (which would indicate the effect on the battery of drawing different current rates is lessened), the SOC dropped.
The value it was at - 1.10 - would tell the monitor the battery had a certain Peukert Factor.
By reducing the value, it is saying the effect of different current rates is not as pronounced as previously stated (with the 1.10 setting). Remembernthr closer to 1.00, the less the effect.
This means that the monitor had been until then, exaggerating the benefit (in terms of greater capacity) of low current rates and conversely, being overly pessimistic on high current rates.
The fact the the SOC went down when I lowered the Peukert Exponent value suggests that my current rate was on average below C20 (the battery capacity you set in the BMV is the makers C20 value), and so the benefit of lower current rates was previously calculated too 'generously'.
If, however, the SOC went UP when I lowered the Peukert value, which is equally possible, then it would be because the average current was ABOVE the C20 rate and the calculation was too pessimistic.

In this particular case, the result makes sense.
By lowering the value I am telling the BMV that the battery is not as affected by current rates as the 1.10 exponent would suggest;
The system has being drawing around the 1.0 to 1.5A range for days and days (and the C20 rate for this bank is 15A)
The SOC reduces as the average current is under C20.

So all the above tie up together. The key thing is IS IT RIGHT?
The logic follows but it doesn't tell you if the actual number is right. And that is down to the operator.
And why I compared the SOC with the Voltage.... This is a brand new setup and needs tuning in. When I looked at the SOC and at the voltage, they didn't really match closely enough... The SOC looked too high. I knew I had a pretty low current draw and concluded that the Peukert Exponent value was not right for this battery bank (the BMV fitted has been on a bunch of different batteries so no surprise it needed settings altered).

If you have the data sheets for your battery, showing the different capacities at different C rates, you can actually plug that info into a Peukert Exponent calculator Victron have on their website and that will give you a number to put into the BMV in fact.
If you don't have the data, then you can do similar to what I did in this case and use some judgement and logic to tweak the number.

 

[HarryW]

Sorry for the confusion using terms such as battery life and efficiency, I just thought that made it simpler in layman’s terms.

the nearer to unity the peukert constant is the lower the ‘loses’ are hence my use of ‘efficiency’ as the descriptor for it to be able to deliver more of its stored energy and hence over a greater time period.

Using the equation in the wiki link. Running the calcs for say a 100aH battery @ 5A load with a k of 1 then running them again for a k of 1.5.

k = 1 gives 20 hours, unsurprisingly
k =1.5 gives just under 9 hours

which is why I raised the question that when you decreased the k value it decreased your reserve capacity whereas my simple logic thought it should have increased it?

I hope that makes more sense...

 

[wildebus]

Sorry for the confusion using terms such as battery life and efficiency, I just thought that made it simpler in layman’s terms.

the nearer to unity the peukert constant is the lower the ‘loses’ are hence my use of ‘efficiency’ as the descriptor for it to be able to deliver more of its stored energy and hence over a greater time period.

Using the equation in the wiki link. Running the calcs for say a 100aH battery @ 5A load with a k of 1 then running them again for a k of 1.5.

k = 1 gives 20 hours, unsurprisingly
k =1.5 gives just under 9 hours

which is why I raised the question that when you decreased the k value it decreased your reserve capacity whereas my simple logic thought it should have increased it?

I hope that makes more sense...

My reply which I added to as you were writing your might explain my thinking further?

 

[HarryW]

My reply which I added to as you were writing your might explain my thinking further?

Yep busy writing before you’d edited your response.

As you say the test conditions are the unknown and your adjustment of the value fits with what you are seeing as the resultant.
Slam dunk empirical evidence, Cannot argue with that.
Thanks for taking the time anyway, btw did you offload that victron 12/12/30 mentioned in a recent B2b thread?

 

[wildebus]

Yep busy writing before you’d edited your response.

As you say the test conditions are the unknown and your adjustment of the value fits with what you are seeing as the resultant.
Slam dunk empirical evidence, Cannot argue with that.
Thanks for taking the time anyway, btw did you offload that victron 12/12/30 mentioned in a recent B2b thread?

I do like the confirm the theories with practical evidence.
Once I get the chance to start using the motorhome and get some heavy loads involved via the interter, I'll probably change the Puekert value in the BMV just to see if my statement about the effect with high current holds water

Out of Interest, talking about the C20 and C100 capacities, the general recommendation from the "official" bodies seem to say you need to look at the C20 figures to see what capacity you will get.
Analysing current pulled from my system on a minute-by minute basis over a camping weekend, using a heater, electric hob, electric water heating, TV, etc, I found my average "C" value was greater (a bigger number) than C100 (I can't recall the exact number without looking at the spreadsheet, but definately more than 100), so I reckon most people will have more actual battery power than they realise!


I have an interested party for the Victron B2B, but if he doesn't take it, I can put you down as first refusal?