Smartshunt inaccurate SOC

Question

question

vanlifer007 asked Jan 31, '21

Smartshunt inaccurate SOC

Hi Guys, Thanks for the great community and support, I have always found this forum to be extremely helpful.

I seem to have an issue with the Smartshunt State of charge in a recent installation.

I have 8x 90AH 12v Yuasa batteries in parallel totalling 720AH.

I set the minimum discharge floor to 60% as my usage is minimal and I would be surprised if I reach that below 80%.

Now the Smartshunt indicate total consumed AH to be -48.1AH and the SOC to be 97%. I think this is inaccurate

Now if I do my math ( I could be ignoring many factors)

according to the set discharge floor of 60% I have 40% usable.

40% of 720AH = 288AH

Expected SOC = 100% -( 48.1/288)x100) = 83.2%

and not 97%

I am confused. can you please help :) below are screeshots and battery specs.

Many thanks

SmartShunt SOC

smartshunt-setting.png (176.4 KiB)

11109-swl2500efr-datasheet-page-001.jpg (261.4 KiB)

smartshunt-history.png (170.5 KiB)

smartshunt-dash.png (120.5 KiB)

2 |3000

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

vanlifer007 answered · Feb 04, 2021 at 05:19 PM

Hi, if anyone is still wondering how to set smartshunt and BMV.

Victron just released a video tackling the questions above. thanks JohnC again.

https://www.youtube.com/watch?v=mEN15Z_S4kE&t=2s

2 |3000

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

Modifications Space

Node Red Space

Answer 1 · 2021-01-31T10:39:49Z

seb71 answered · Jan 31, 2021 at 10:39 AM

The "Discharge floor" setting from the SmartShunt has nothing to do with the SOC value and also does not stop the battery discharge.

It is only used to estimate the "time remaining" (time to go).

Regarding the inaccurate SOC value:
You have to start from a fully charged battery and then set the SOC to 100%.

Also you have to tweak the "automatic synchronization to 100% SOC" settings (this is controlled by: charged voltage, tail current and charged detection time settings). Observe and adjust. Repeat untill you get satisfactory results.

3

2 |3000

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

vanlifer007 commented · Jan 31, 2021 at 10:58 AM

Thanks for your input.

I understand that the SOC value will not stop the discharge, however it will triger alarms, relays...

it is just that the actual draw doesnt reflect the SOC. be it a new install, I will keep a close eye.

0 ·

seb71 vanlifer007 commented · Jan 31, 2021 at 11:01 AM

I did not talk about the SOC value there. I talked about the "Discharge floor" value.

0 ·

vanlifer007 seb71 commented · Jan 31, 2021 at 11:09 AM

agreed, not the SOC nor the "dischagre floor" will stop discharging.

if the SOC takes into account the full AH battery capacity (720AH) and not the set discharge foor (40% in my case)

48AH to 720 AH is 6.6% the Shunt should indicate 93% and not 97%

0 ·

Answer 2 · 2021-01-31T12:45:10Z

JohnC answered · Jan 31, 2021 at 12:45 PM

Hi @Vanlifer007

With a Peukert Exponent above unity, and drawing other than the C20 discharge rate, then the SOC cannot match the Ah drawn. That's by definition, it's impossible.

I ran your batt specs through the Victron Peukert Calculator. It bobbed with 1.06, and you could change the existing 1.25 to that. You'll get somewhat closer, but it still won't match as you expect. It's not designed to. And Ah isn't a real measure of energy anyway, it lacks V.

Some other stuff:

Your batts are actually rated 93.6Ah ea at C20, and you should use that.

95% Charge Efficiency is likely too optimistic, and reality is probably closer to 90%.

The 4% Tail is probably way too high. I can't say where that should be, but I use 0.5%. Too high will cause premature sync, and again overestimate SOC.

8x batts in parallel isn't good for longevity. Imbalance the issue there, particularly as they age. Take the time occasionally after a hard charge session to feel them for temp differences, and if you can feel it you have a problem. Hopefully your wiring is good, and doesn't initiate imbalance.

3

2 |3000

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

vanlifer007 commented · Jan 31, 2021 at 11:10 PM

Hi @JohnC, as always I i patiently await your response and expertise. Thank you very much indeed for your time and input.

you commented on my setup last year 8x parallel. and you were right again, the MPPT can not get them to absorption, but again UK weather what to expect :).

I would like to think that my wiring is good, I followed a balanced wiring schematic where each battery have same number of knots . I know that unless there is lab conditions it would be hard to achieve equal draw, but so far the batteries seems fine (touching wood) no difference in temp and when I disconnected them recently to move them they all measured the same voltage...

Now if I go back to my question, I think I made mistake attaching the wrong battery spec sheet, I followed your advice and have been monitoring usage this afternoon, I can see some progress where the SOC dropped to 95%.

as you can see from the history I have just installed the smartshunt. I havent synchronised the shunt manually, as I am not sure when is 100%. and since the batteries havent seen float since the summer I have been just topping them up from driving around, I can see initially 70A coming from the alternator which drops to 30A and not below whilst the voltage go up to 13.9-14v. is this 100%? that is what the shunt think.(an Orion-tr is on my bucket list although no VE.Direct),

Can you kindly review the correct spec sheet attached below and advise if the Peukert , Charge Efficiency and Tail are same as previously advised?

Can I be sneaky since you are here and get you to take a look at my MPPT set up and see if everything is correct or close?

I have everything connected to a PI running venus OS, I have some reading to do about DVCC, is this a better route to get the best of everything?

0 ·

yuasa-swl2500fr-page-001.jpg (312.5 KiB)

mppt-history.png (126.8 KiB)

mppt-setting-1.png (154.9 KiB)

mppt-setting-2.png (125.0 KiB)

JohnC ♦ vanlifer007 commented · Feb 02, 2021 at 11:39 AM

@Vanlifer007

I had to find another specsheet to find the C-ratings, but they were slightly different, Still worked out 1.06 Peukert, but 92.4Ah at C20.

Guessing your mppt is a /50 unit, but your production is tiny, and that's the reason you're not reaching Absorb.

The alternator's 30A means a ~4% Tail, and if it's not going lower, then that should be 100% charge. Surprisingly high, but if that's what it is it's way higher than what your solar can do.

Note the Tail setting in the mppt is in A (not %), and is to drop the mppt out of Abs to Float. DVCC is great for handling this, but the 2A you have there should be up above the 30A you estimate as the 'tail'. DVCC can also give you a common V and Temp for all your kit to work from, and if you have loads running also from the mppt it can also use the Shunt A for it to decide to drop from Abs (batt A rather than mppt A). DVCC is worth doing.

The 14.1 Abs setting is a little lower than the cyclic spec, but you mustn't be cycling much, so no real issue. You don't say how much pv you have, but not much happening there anyway..

2 ·

vanlifer007 JohnC ♦ commented · Feb 02, 2021 at 01:19 PM

Hi @JohnC.

cheers for the answer and the Peukert I couldnt find c20 anywhere.

I thought of the 4% tail when the amps did not drop below 30A, I will have to take the van for a long drive and monitor this closely.

Yes it is 100/50 mppt Not much PV going on, 310w AEG mono panel. not much roof space in the campervan I am afraid:). I am planning to pump this by adding a panel that will slide out and act as an awning when stationary or add series of folding panels. but maybe will have to downgrade the mppt or add a seperate one as the small panels will downgrade the output :)

30A tail seems a lot. here is a snapshot from last year, in the summer I used to see all 3 cycles of charging and sometime the batteries never drop below 13.10.

Take a look at Grafana dashbording from today(credit to Marks). I am having a bit of fun with it, along with home assistant, see if I can automate the dump load for the water heater ( when there is sun) but that is a different story :)

I really appreciate your answeres John, you are one of the legends here :).

0 ·

aeg-solar-panel.jpg (28.3 KiB)

mppt-2020.png (138.4 KiB)

grafana.jpg (200.4 KiB)

Answer 3 · 2022-10-26T20:15:49Z

Harold Lopez answered · Oct 26, 2022 at 08:15 PM

Hi @JohnC , I 've got a question about the peukert value that you recommend.

I thought that a peukert near 1.05 is just for Lithium, and Victron says for lead acid 1.25.

I have a similar issue with a 510AH, 48V system, where the voltage drop to 47.5V and the shunt still shows 87% SOC. My peukert is now in 1.15, and tail current in 4%.

thanks in advance for your time and patience.

2

2 |3000

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

JohnC ♦ commented · Oct 27, 2022 at 03:11 AM

Hi @Slopz01

That default 1.25 isn't necessarily wrong, but it may be more accurately applicable to high discharge rates of like C5 or C10, which most of us don't use. Usage like that may well deserve the SOC 'penalty' it applies. Typical usage might be closer to C50, drawing down batts overnight, so a high Peukert then will give a high 'credit' to the SOC, and can lead to an overly optimistic SOC. So I start with a Peukert from the Victron calculator (using C20 to C100 batt specs = 1.16), then reduced it to 1.10 to err on the safe side.

Critics might say that's rough. Even Mr Peukert, whose algorithm is empirical anyway, admitted that temperature isn't considered. So I agree it's rough. But you can get awfully close if you get to know your batts and experiment.

Like you, I don't like to see my Pb's go below 48.0V. But they can, even at SOC >80%. A little load and winter batt temps can easily drop them below 48. With confidence in your SOC you can shrug that off.

Your 4% Tail is too high for my liking, but if your Charged V is up around Absorb charge level, then you'd sync to 100% at maybe 98% true. Close enough if you're happy with that and understand it.

With the two factors Charge Efficiency and Peukert, the aim is to adjust them to return to 100% SOC on a daily basis, naturally, at the right time. Gotta be careful of what your sync is doing though, that's a step that can foul up your careful work.

Note too that VictronConnect can show in real time what an adjustment to Peukert will do to the SOC if the batts have discharged some. And it's reversible, a good tool to experiment with..

1 ·

Harold Lopez JohnC ♦ commented · Oct 29, 2022 at 06:18 PM

@JohnC Thanks a lot.

I'll follow your advice.

0 ·

question