wattmatters avatar image
wattmatters asked

Smart Shunt SOC readings at low discharge rates

Am I right that due to the Peukert formula the Smart Shunt reported state of charge (SOC) for low discharge rates will fall more slowly than the reported Ah discharged would suggest?

I added a SmartShunt 500A/50mV to keep better tabs on my sealed lead acid battery bank.

I've noted most nights SOC does not drop as much as Ah discharged suggests it should. Overnight discharge rate is low (~2 A from a 380 Ah battery bank), so in the vicinity of C200 rather than C20.

e.g. I see Ah suggest SOC should have dropped 6-7% when the SOC indicator drops by only 3-4%.

Current settings:


Looking at my battery's specifications the Peukert Exponent works out to 1.13, somewhat lower than the default value of 1.25. These are telco/data centre backup SLA batteries and a Peukert value of 1.13 seems to gel (pun intended) with such batteries (not accounting for any temperature or age factors).

Might try a Peukert exponent of 1.13 and see what effect that has.

As an aside, working out the Peukert exponent made me recheck my battery's specifications, and the rated Ah is at the C10 rate with discharge down to 1.80 V/cell.

The C20 Ah rating is 11.5% higher than that.

So based on a C20 rating my battery bank would be 424 Ah, not 380 Ah. Interesting.

img-3627.png (184.3 KiB)
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

2 Answers
kevgermany avatar image
kevgermany answered ·

Peukert describes battery behaviour, doesn't control it.

If you're constantly operating in the C200 range, yes increase the capacity in the shunt settings. And yes change peukert to 1.13. Maybe check with the supplier for more appropriate values at C200.

1 comment
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Thanks, yes I understand Peukert is just a model of battery behaviour. I was more just getting my head around how the SmartShunt was actually calculating SOC. The manual and guides never actually say.

If it's just a straight exponential function and the exponent is based on C20 as a baseline value then I would expect C rates < C20 will see SOC fall more slowly than Ah discharge value might suggest.

I think what I'm getting at is the model probably has a domain of validity and a discharge rate of C200 is outside that domain.

I changed the exponent in the settings and will see how that goes.

C200 overnight it kinda normal as this bank is mostly designed for outage backup. But the plan is to add capacity and then begin to use it for more regular household loads.

0 Likes 0 ·
JohnC avatar image
JohnC answered ·

Hi @wattmatters

Victron rarely publish their algorithms, I guess to deter cloning. Peukert indeed will compensate for a low discharge rate, and effectively call the Ah out (towards the calculation of SOC) as from a higher capacity battery. So there's no point entering a C200 batt capacity when this is already being compensated for. Just use the expected C20 rate, and calculate the Peukert from the batt maker's figures on the C200 side of C20.

Too high a Peukert will overestimate SOC, and I prefer to run with the safe side, ie. a lower Peukert. My batt specs say 1.16, which I use for winter, but drop it to 1.10 in summer. Mr Peukert didn't quantify the effect of Temp, but I did, albeit coarsely.. :) I also need to drop Charge Efficiency from 93% to 90% in summer. Temp has a huge effect.

If you don't mind me saying, you may have other corrections to make too in your settings. Your CV is too low, your Tail too high, the CDT maybe too low. That's aside from the Peukert and CE (too high?).

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Hi John, thanks for your comment/suggestions. I only just installed the Smartshunt the other day and the settings shown in the pic are Victron's default/recommended settings for the battery type, aside from the battery capacity I entered.

I figured over time I would work out what settings made more sense for my set up.

I changed the Peukert exponent yesterday to 1.13 and overnight the SOC has fallen more than the night before, and a bit closer to what I would expect based on the Ah discharged.

If the shunt expects a C20 capacity value (rather than the C10 value the batteries specify), then I can change that to the C20 specification. Depends on the end voltage used as my batteries gives specs for discharge to 1.75, 1.80 and 1.85 V/cell. The 1.80 V C20 rating for my bank would be 424 Ah (although they are not new batteries so who really knows?).

I agree with your comments about Charge efficiency, Tail current and CDT. I'm not as clear on the CV though.

Manual says CV should be 0.2-0.3 V below float voltage. They don't however say if that's per 12 V battery or for the battery total in series. Mine float at 54.0 V. If they mean 0.2-0.3 V per 12 V battery then the recommended 52.8 V makes sense, but if it's meant to be for the full bank then 53.7-53.8 V would make more sense. Not sure how this setting is going to do much as the battery is either charging or discharging, rarely resting.

Charge efficiency did seem oddly high to me. I might ease that back.

Tail, yes that seems high as well. I know after a decent float time my batteries settle into a tail current at ~1% of capacity rating. I guess I can drop that back to say 2%.

I'll increase the CDT to 10 minutes (longer?).

Of course these parameter settings will get tossed out the window when I start my experiment with a using a hybrid battery set up - I'll be adding some LiFePO4 to the mix and it will manage the bulk of the daily cycling when I start using the battery for managing general household loads, while the lead battery will just sit there on float most of the time, ready for backup duties.

It will be fun to see how it performs.

0 Likes 0 ·

In VictronConnect you can actually change the Peukert and watch the effect it has. In a discharge scenario of course. Then change it back. Cool little feature.

CV should be out of range of the temp-compensated charge setpoint. 0.1V should be fine if it's steady. If you're using a low Float, then syncing to it gets a little tricky.

CDT increase is to allow for the 'passing-cloud' effect, where A drops instantly, but V sluggish to follow. 3 min catches me occasionally, but 5 min ok.

0 Likes 0 ·