Scheduler for INDEVOLT home battery

[Artenverho]

I have made a battery scheduler for my home battery from INDEVOLT. It is based on a python plugin I wrote to obtain most the sensor data for the battery via a local HTTP api (one of the nice features of this particular battery!). Then using 6 dzvents scripts:

• Quarterly prices data is obtained from Enever, or directly from TIBBER. it is saved as a JSON to two user variable

• Hourly solar prediction is made based on data from NED.nl and adjusted to the position and orientation of my PV panels. saved as JSON to user variable

• A script tracking the value in EUR/kwh that is present in the battery based on how much is charge during the current price slot. updates a dummy sensor

• The battery scheduler itself: this script responsible for gathering all the above data and making a schedule for the most economical charging and discharging of the battery. It takes into account the times during the day that solar energy is available and discharging only occurs to my own appliances (i do not want to export the energy back to the grid). When done the scheduler provide the schedule and a summary to user variables.

• Battery scheduler execute script that reads the schedule JSON and sets a dummy selector switch to charge, PVcharge, discharge or idle.

• Battery controller script that reads the above mentioned selector switch and sends the proper commands to the battery.

This all runs automatically without my input. I wanted to be able to monitor it and, if I desired, manually fine tune the schedule. That is why I modified the energy dashboard to include a battery scheduler section:


This small dashboard allows me to quickly view a summary of the latest schedule, manually adjust charging and discharging slots and view the prices at a glance.

I am hesitating however if it would be worthwhile to share everything as is. Currently, the scripts are very much tailored for my own setup with the INDEVOLT battery so I am not sure how well somebody else could implement it. Curious about any opinions on this .

[jake]

De INDEVOLT is een zeer prettig geprijsd product. Ik heb chatGPT er aan laten rekenen en met de huidige saldeertarieven (21ct kopen en 15ct teruglever vergoeding) zou het bij 1 cycle per dag een tvt van 6 jaar hebben bij een 2kWh variant.

Kan je zelf bepalen met hoeveel vermogen er geladen en ontladen wordt?
Hoe gaat het met een 3-fase aansluiting, aangezien je uiteraard maar op 1 stopcontact invoert.

Overigens vraag ik mij af of een gewoon stopcontact het wel zo lekker doet met een langdurige belasting van 2400W. Of dat toch niet warm gaat worden

[Artenverho]

Yes the API allows you to set discharging and charging at any power setting between 0 and 2400watt. I personally only discharge for own use. this can also be engaged through the API. The battery uses its own P1 meter to make sure it delivers exactly the power needed to keep the meter at 0. When it comes the 3 phases the meter does not really keep track how much it exported/imported for each individual phase. so yeah technically you are supplying the grid on one phase while likely consuming on the other ones.

My system is connected to a socket that has dedicated wiring to its own circuit breaker in my breaker panel. I haven't noticed any serious heat from the plug/socket to be honest even after 2.5h hours of charging (at which the battery is full). It might be an issue if you do it for longer? in any case discharging rarely spike above 1000w for me as I do not discharge to the grid.

Alternative would be to cut the power cord and wire it permanently into a connector box. It might do that but I am actually not really convinced it will be safer...

[waltervl]

@jake, where are you going to get 5 cents of delta between usage and return? Are you in Belgium? As in the Netherlands the delta will be 0.5 cents of even less.
So Return On Investment time will be much longer.

[jake]

@jake, where are you going to get 5 cents of delta between usage and return? Are you in Belgium? As in the Netherlands the delta will be 0.5 cents of even less.
So Return On Investment time will be much longer.

I just got a new contract with green choice for 1 year. Until the end of this year, we will be able to 'salderen' (to subtract supply to the grid from demand from the grid) and that results roughly in a delta of 7ct.
For next year nobody knows what the energy suppliers will pay for supplied energy. If it even less than the 7ct, the ROI period will be shorter.

1 more question, the P1 is already connected to Domoticz. Can I still communicate that data with the BMS?

With the API you control how much power you allow in both directions of the battery?

[waltervl]

You can get a quote for a 2 year contract, the you will probably see it will be way less then 7 cents.
A home battery in the Netherlands next year will hardly pay off, it will only help with using more solar energy directly. But that will give long ROI times as the energy tarifs are low.

[Artenverho]

You can get a quote for a 2 year contract, the you will probably see it will be way less then 7 cents.
A home battery in the Netherlands next year will hardly pay off, it will only help with using more solar energy directly. But that will give long ROI times as the energy tarifs are low.

Hmm I am not so sure.. if you have a dynamic contract (like I do) next year will actually allow you to have a decent difference in energy price daily. You can then store a lot of your solar energy during the day and discharge during the evening and night. I will be speculative how much it will actually be (prices obviously fluctuate) but given that next year you do not get any tax back for your own generated electricity there at least a ~12cent gap... during the past year on typical sunny day the gap would be about 8-10 cents between solar peak and evening, so combined this is a considerable spread.

Important note: this assumes that you use all the energy use store yourself! therefor it is also crucial you don't install too much capacity for it to be economical. I have about 6kwh (1400euros incl. btw) and this should be about the amount i can justify according to my own consumption. ROI is difficult to predict but pessimistically it will be about 8 years. optimistically 6... we will see .

ohh and for this year I expect a lot less advantage as it will often be more economical to feed the solar generated electricity back to the grid.

[jake]

I have about 6kwh (1400euros incl. btw).

Last time I checked, the price is about 600 euro/2kWh.

Still relatively cheap, usually I see prices of 5000-6000 Euro/10kWh. What would cause the price difference?

For years I have a 'software battery' running in Domoticz:
https://github.com/jakenl/domoticz_solarbattery

I chose a capacity of 5kWh and I see that in summer times it quite often is full and not emptied and in the winter it is never full and most of the time it is empty.

2, or maybe 4kWh seems to be plenty

[Artenverho]

Last time I checked, the price is about 600 euro/2kWh.
Still relatively cheap, usually I see prices of 5000-6000 Euro/10kWh. What would cause the price difference?

For years I have a 'software battery' running in Domoticz:
https://github.com/jakenl/domoticz_solarbattery

I chose a capacity of 5kWh and I see that in summer times it quite often is full and not emptied and in the winter it is never full and most of the time it is empty.

2, or maybe 4kWh seems to be plenty

I think the price differences are mainly because the more expensive system can sustain much higher charge/discharge loads and sometimes also work as a off grid backup. the INDEVOLT can do the latter but only for one socket on the device.

6kwh will also be plenty for me indeed and yes during the winter I indeed almost never fully charge/discharge it in 24h. It still is useful though as I occasionally charge from solar and in other cases when price peaks occur I charge from the grid at night and discharge during the peak hours in the evening.

In the summer I suspect I will use the 6kwh fully as I have a floor cooling system that uses about 350kwh constantly when it is above 25 degrees outside. so in the evening/night/early morning I think I will deplete the battery. We will see though. It would have been nice to test this last summer with your software battery had I known about it. looks nice!

[BarryT]

I have made a battery scheduler for my home battery from INDEVOLT. It is based on a python plugin I wrote to obtain most the sensor data for the battery via a local HTTP api (one of the nice features of this particular battery!). Then using 6 dzvents scripts:

• Quarterly prices data is obtained from Enever, or directly from TIBBER. it is saved as a JSON to two user variable

• Hourly solar prediction is made based on data from NED.nl and adjusted to the position and orientation of my PV panels. saved as JSON to user variable

• A script tracking the value in EUR/kwh that is present in the battery based on how much is charge during the current price slot. updates a dummy sensor

• The battery scheduler itself: this script responsible for gathering all the above data and making a schedule for the most economical charging and discharging of the battery. It takes into account the times during the day that solar energy is available and discharging only occurs to my own appliances (i do not want to export the energy back to the grid). When done the scheduler provide the schedule and a summary to user variables.

• Battery scheduler execute script that reads the schedule JSON and sets a dummy selector switch to charge, PVcharge, discharge or idle.

• Battery controller script that reads the above mentioned selector switch and sends the proper commands to the battery.

This all runs automatically without my input. I wanted to be able to monitor it and, if I desired, manually fine tune the schedule. That is why I modified the energy dashboard to include a battery scheduler section:


This small dashboard allows me to quickly view a summary of the latest schedule, manually adjust charging and discharging slots and view the prices at a glance.

I am hesitating however if it would be worthwhile to share everything as is. Currently, the scripts are very much tailored for my own setup with the INDEVOLT battery so I am not sure how well somebody else could implement it. Curious about any opinions on this .

Would you share this? Very interesting..
Thanks

[willemd]

You can get a quote for a 2 year contract, the you will probably see it will be way less then 7 cents.
A home battery in the Netherlands next year will hardly pay off, it will only help with using more solar energy directly. But that will give long ROI times as the energy tarifs are low.

Low tariffs is not a problem,since the difference between return price and usage price will be approx 12 cents (based on current energy tax) and usually more since solar energy that would normally be returned during the day at 0 to 5 cents might prevent purchase during the night at 30 cents or so.

I simulated with real 2025 data that I will have to pay approx 350 extra in 2027 because end of "saldering" = netting of usage and return. A 600 euro 2.1 kwH home battery can reduce that amount for me with 120 euro, so payback would be 5 years. This is achieved by storing solar energy instead of returning it to the grid and using it later (during nighttime).

Double capacity would save me another 80 euro, so less gain per kw or euro.

If have made a battery planning program that uses lineair programming optimisation to maximise financial return. This year by taking advantage of differences between prices during the day, next year by using solar energy to prevent power import during the night. It takes day ahead prices, solar production forecast, usage forecast, battery specs (from external sources and domoticz)

https://github.com/WillemD61/battery-planning

[BarryT]

Overigens vraag ik mij af of een gewoon stopcontact het wel zo lekker doet met een langdurige belasting van 2400W. Of dat toch niet warm gaat worden

I have 30 kWh of IndeVolt storage here, divided over 3 packs of 10 kW each.
It is NOT permitted to set the output power to 2400W when the unit is connected to a standard wall socket. This is also clearly stated everywhere in the documentation.
First of all, you absolutely should not want to do this, and secondly, your insurance company may create serious issues if it turns out the batteries were configured for 2400W while connected to a socket circuit that has additional loads connected to it. Keep that in mind.
The maximum output power of a battery may never exceed 800W when connected to a standard wall outlet (WCD), unlike a fixed installation with its own dedicated circuit, where up to 3680W is allowed.

Ps, for a lot of techical talk about this and batteries you could take a look on the tweakers forum.

[willemd]

Overigens vraag ik mij af of een gewoon stopcontact het wel zo lekker doet met een langdurige belasting van 2400W. Of dat toch niet warm gaat worden

Ik heb hier 30kwh aan opslag van IndeVolt verdeeld over 3 packs van 10kw.
Het is NIET toegestaan om de 2400W aan vermogen in te stellen wanneer deze in een "algemeen" stopcontact is aangesloten, dit staat ook duidelijk overal vermeld.
Ten eerste moet je dat absoluut niet willen, en ten tweede, de verzekering kan heel moeilijk doen wanneer blijkt dat het vermogen van de batterijen staat ingesteld op 2400W en deze zijn aangesloten op een stopcontact waar meer op zit. Houd daar wel rekening mee.
Het maximale vermogen van een batterij mag nooit hoger zijn dan 800W wanneer deze is aangesloten op een wcd, in tegenstelling tot 3680W op een vaste aansluiting met een eigen groep.

English please in this forum. Please use google translate if needed.

[BarryT]

Changed. Excuse me