Lux ac battery controller

Petriix

I think you might be forgetting to compound the losses for both charging and discharging. In addition, the whole process will be less efficient when charging and discharging at lower rates - say meeting your overnight base-load.

1kWh x 0.95 (inverter losses) x 0.9 (battery losses) is 855Wh stored in the battery. 855Wh x 0.9 x 0.95 is 733Wh; ~73% of your original energy.

ianatkin

Petriix said:

I think you might be forgetting to compound the losses for both charging and discharging. In addition, the whole process will be less efficient when charging and discharging at lower rates - say meeting your overnight base-load.

1kWh x 0.95 (inverter losses) x 0.9 (battery losses) is 855Wh stored in the battery. 855Wh x 0.9 x 0.95 is 733Wh; ~73% of your original energy.

I knew I needed to compound the round trip losses on the inverter side but I didn't appreciate losses on the battery side would be as high as 10% each way, conveniently there isn't an efficiency stated on the battery datasheet as far as I can see...

This has all led me to realise that, rather than setting the shortest possible AC charge time on the inverter to add whatever I need, I should have the AC charge timer set for the entire five hour cheap period on my Go Faster tariff - better to use 5.5p grid power directly than let the battery get used during the cheap period and then have to put it back in - that would effectively cost more like 7.3p accounting for all the losses!

QrizB

ianatkin said:

This has all led me to realise that, rather than setting the shortest possible AC charge time on the inverter to add whatever I need, I should have the AC charge timer set for the entire five hour cheap period on my Go Faster tariff - better to use 5.5p grid power directly than let the battery get used during the cheap period and then have to put it back in - that would effectively cost more like 7.3p accounting for all the losses!

I've gor a Sofar and I'm on Go, but do something similar. I tell the Sofar to start charging at 0040, and to stop at 0420. Usually it finishes in a couple of hours but it stops the Sofar trying to service my background loads until the cheap Go hours end.

(I use 0040 and 0420 in case the Sofar clock drifts.)

billy_fixit

 I have Octopus Go (not Go faster) so I get 4 hrs cheap rate. I set my Lux controller to charge ALL the time from 00:35 -> 04:25. This stops charging the batteries when full (assuming enough charge rate to get them full in <= 4hrs which can always be done). but stops the batteries discharging at all during the above charging period, so never need to put charge back in again during that 4hr period. Then charging period ends at 04:25 and the batteries take over from then on (as I have the LUX configured)

So it seems that ~~75% efficiency might be about the real-world practical experience of folks.
Pylontech batteries are an obvious alternative to the AOBOET units (at least with the otherwise pretty good LUX box) and I think saw a recent post saying that they seem to exhibit ~~75% efficiency too.

Perhaps there is nothing to choose between brands in terms of real-world efficiency but, I'd like to ask folks to comment on the SOC tracking/reporting of the Pylontech batteries (via the LUX controller?). Do folks find the SOC reporting to be reasonable for Pylontech? Certainly the AOBOET batteries do not seem to report SOC very well and it can vary wildly with very small charge/discharge events. Do folks also see this with Pylontech or is the SOC monitoring observed to be better?

Thanks for the recent comments.

newbuilder

Can I ask what happens tonight with Lux times and moving to GMT?

EVandPV

billy_fixit said:

 I have Octopus Go (not Go faster) so I get 4 hrs cheap rate. I set my Lux controller to charge ALL the time from 00:35 -> 04:25. This stops charging the batteries when full (assuming enough charge rate to get them full in <= 4hrs which can always be done). but stops the batteries discharging at all during the above charging period, so never need to put charge back in again during that 4hr period. Then charging period ends at 04:25 and the batteries take over from then on (as I have the LUX configured)

So it seems that ~~75% efficiency might be about the real-world practical experience of folks.
Pylontech batteries are an obvious alternative to the AOBOET units (at least with the otherwise pretty good LUX box) and I think saw a recent post saying that they seem to exhibit ~~75% efficiency too.

Perhaps there is nothing to choose between brands in terms of real-world efficiency but, I'd like to ask folks to comment on the SOC tracking/reporting of the Pylontech batteries (via the LUX controller?). Do folks find the SOC reporting to be reasonable for Pylontech? Certainly the AOBOET batteries do not seem to report SOC very well and it can vary wildly with very small charge/discharge events. Do folks also see this with Pylontech or is the SOC monitoring observed to be better?

Thanks for the recent comments.

No issues here with the Pylontechs, the SOC is as expected with the Lux or when connected directly to the batteries via serial cable and using the Batteryview app.

EVandPV

newbuilder said:

Can I ask what happens tonight with Lux times and moving to GMT?

Can't remember ever having to do anything. I'm sure it all just takes care of itself regarding the clocks changing.

ianatkin

I just logged into the app using local connect and the time hadn't changed automatically, I've just changed it manually. Will keep an eye on it in case it decides to take another hour off itself later.

Solarchaser

Taking battery discharged over battery charged my efficiency is 82% with around 10kwh useable pylontech on each lux. 

Pylontech tend to be totally linear up to about 93-95% when they decide they are full pretty quickly 

EVandPV

Solarchaser said:

Taking battery discharged over battery charged my efficiency is 82% with around 10kwh useable pylontech on each lux. 

Using that method, I get 83% efficiency.