Did you know that most alfen chargers can be controlled by evcc and precisely utilize the PV surplus? And evcc has an openhab binding so you can control everything from openhab.
Do I understand this correctly that these hardware components you mentioned did the magic so that the power that was supplied to the boiler was dynamic i.e. not simply ON/OFF at 2 kW?
Yes, I know evcc but haven’t checked how it works with Alfen charger. I have also looked at Alfen’s Modbus protocol but there was something which I wasn’t quite happy about. In principle Modbus is rather straightforward and I have several devices which communicate with my OH5. I’ll have a look at evcc.
Yes, basically I just have the SSR which is controlled by Ouman. I use this Ouman for many other purposes so there is perhaps no point to get this controller only for controlling the boiler. Have a look at the links in my post. Perhaps there is a cheaper solution.
Yes, the SSR provides an adjustable AC voltage for my boiler, not just on/off. If I remember correctly you also have a Nibe boiler. Which model it is? How many phases it has?
Shelly ev charger. You’d even help in making it compatible with the Shelly addon 
Yes, it’s a Nibe Haato 300, the manual shows the two connection options:
- 3 kW with three phases (this is how it is currently connected, my relqy + contactor cuts / allows all 3 phases)
- 2 kW using a single phase
According to the electrician I talked to, I could achieve the 1 kW / 2 kW / 3 kW steps simply by controlling the three phases individually.
Yes, you can control the 3-phase system by controlling each phase separately but the step is 1kW which means that you can’t use the excess PV power effectively. The 1-phase system is easier to control with a PID controller and a SSR. Your boiler has 300 l of water so 2kW heating power is not perhaps quite sufficient.
You could think about getting batteries because then you wouldn’t need to worry about optimizing your boiler and ev car. You could also run your heat pump with normal settings so no need to overheat your house during cheap hours.
Thanks @jlikonen ! I think I now understand conceptually what the SSR is doing together with the PID controller.
Our ground source heat pump is heating the cold water in its 180 liter tank and then supplies this already heated water to the 300 liter boiler which is this 3 kW Nibe Haato that I habe been talking about.
Since the water is already pre-heated, this boiler only consumes like 5-6 kWh per day i.e. 1.5 - 2 hours with 3 kW power.
If I change the EV charger to a model which can dynamically adjust the charging power, I can re-use the contactor that is controlling our current charging station so then I would only need 1 more contactor to achieve the 1 kW / 2 kW / 3 kW steps for the boiler. That part I fully understand, and the limited understanding on this SSR and PID concept makes me hesitate a bit. I would probably understand it sufficiently by reading and asking questions but it sounds a bit more complex than just controlling each phase separately.
Even though I don’t have experience with PV systems yet, I’m quite optimistic that these 1 kW steps would make quite a big difference compared to 3 kW on/off approach so if we decide to go on with PV, then I’ll probably start with that.
Big enough batteries would of course solve this whole dilemma and would also help during winter to cut the most expensive peaks, but I’ afraid that the “minimize the money paid to Caruna, no matter how much it costs” might have a quite small WAF 
There are also 3 phase SSRs and you can get pretty reliable manufacturers for a price comparble to a contactors. The more expensive part would probably be the PID controller. SSR is in general gentler to the heater element than a contactor, and it will work for a simple boiler, however chopping the sinusoide might cause problem if the boiler has some advanced functionality like for example legionella elomination cycles.
Regarding the solar generation, I have 11kW pv panels, that generate up to 60kWh in the summer and rarely more than 10kWh in the darkest days of the winter (occasionally I get as low as 1kwh per day) and this is the periode when my heat pumps use the most.
I also did some consideration to install batteries, but after doing some calculations I am not convinced that the payback is fast enough.
Lets say that I generate 10kWh surplus per day for 200 days a year and if I take in account the cost difference for what I pay for that electricity decreased by the amount of what I get paid for the sold power, it comes out that the payback time is 10+ years including the losses. After those 10 years you can also expect some failures, so my decision is not to proceed with it.
What I did instead is to heat my hot water boiler with the excess and try to charge the car when it is connected overday, but as in your case that is almost only in the weekend. During the weekdays I charge it overnight on cheaper tariff.
On WAF and strategic questions…
I got my inverter several years ago. It is capable to run as an island during power outages. Batteries were a lot more expensive back then, return on investment was negative in fact.
But that capability actually was the reason my wife opted to nonetheless go for it.
The conditions have changed a lot since, I’m now even making money with it, my wife’s happy and so am I that I got them.
Size matters - so get a small one but make sure it’s always working to make your wife happy
Absolutely… you can buy a fairly decent hybrid inverter for 200 euro and 625 watt panels for 90…. 5kwh batteries for 800 euros..
It’s crazy.
Okay so let’s now bring batteries to the table. The benefit side are quite clear
- it will act as a buffer so that I don’t need to sell surplus energy back to grid with close-to-zero price just to purchase it back later and pay the grid tariff and taxes
- during dark winter time, I can load the battery when it’s cheap during the night and consume it during more expensive daytime hours
- (In some cases it might even be beneficial to sell back to grid from the battery but I won’t go there now for the sake of simplicity)
What would I need to consider about the battery stuff?
- I would need the physical batteries of course
- And presumably a separate inverter for the batteries i.e. PV system would have its own inverter and battery pack would have its own? Edit: I learnt that the term “Hybrid Inverter” means an inverter which can handle both.
- Talking about inverters (both PV and batteries), are they able to supply power to all three phases so that car charging, water boiler or the ground source heat pump can use all 3 phases?
- I guess there are different kinds of safety considerations for both the batteries and inverters?
There’s of course all the options available out there. But there’s no advantages in getting separate or 1-phase inverters. You would only get one in edge scenarios or if you already have a PV-only inverter in operation. Long story short, get a 3 phase hybrid inverter right away.
It very much depends on selecting a good inverter and getting an installer that provides and supports your specific brand. Selecting the right experienced installer is even more important if you look into advanced features like island operation. Either way, he/she should also take care of safety considerations. I would never go DIY. Insurances are very picky when it comes to regulating high-value damages related to electrics or fire.
And like I said, you can get a modular battery system.
Not sure about others, but BYD is the most well known battery vendor, supported by most inverters. They have a stackable system so if you want to bet on prices to fall, you could get a minimum system together with the inverter and add more battery blocks yourself later.
Just to note it: You cannot add new blocks after two or three years (IIRC) depending on the degradation of the existing batteries.
In addition to adding new blocks to a single tower, many inverters also support horizontal scaling by putting several towers besides each other.
Have reviewed the 2026 battery inspection paper from HTW Berlin. Thanks for this link. The paper is available in German and in English.
There is indeed one german company near Ulm which provides outstanding storage systems 220V to battery and back. These are independend from already existing combinations of PV and inverter. They have focussed on high efficiencies of >96-98% for small power, lets say in the range of 300W to 1kW. This is likely the standard night consumption of a normal house without an EV. But it is also the weak point of most 5 or 10kW hybrid inverters which have their best efficiency in the higher kW range.
Comming back to my efficiency point, it is very important to look for the target discharge power range if a battery system is selected.
We have the Fronius Ohmpilot in our system for controlling the hot water heater. The hot water is heated with a 3 phase 3 kW heating element. The Ohmpilot connects power to the element with relays for two phases each 1 kW but the third one is stepless between about 100W to full 1 kW. This arrangement makes it possible to use solar surplus between 100 W to 3 kW to heat the hot water. During the day there are many occasions where there is a small surplus and with this control arrangement this surplus can be precisely used. Ohmpilot has a small control delay so the setting doesn’t change for every small cloud that affects the solar radiation.
Propably there are devices (Shelly?) out there that you could control with some code to get the same functionality as Ohmpilot has?
Regarding inverters and batteries our Fronius GEN24 inverter with a BYD 10,24kW is able to supply power to all our 3 phase loads if needed. To my understanding the Fronius GEN24 hybrid inverters are one of the few that can provide 3kW 1 phase power from the panels and battery in case of grid failure. It has this PV output that is automatically activated after 30 seconds after grid failure is detected. This is a very good feature to prevent fridge and freezer meltdown, run your water well pump and other small power appliances. Rewiring some electricity circuits could automate things in case of a grid failure but for the time being we handle the situation manually using a couple of extension cables as we are both retired and almost all the time at home.
Ultimately - well not so much. Sure if you can get one that’s efficient for all of your power levels incl. nightly ones, please do. However, if you need to choose between this one and another one that has better features or price, I’d not rate efficiency a high priority. Remember it’s fairly low wattages so the absolute amount of power being lost doesn’t add up to big figures.
The other factors are more important. Here’s another recent thread on this:
X2
Thank you for saying this. You can’t really have both ways, you either have good efficiency at your idle values, or you have good efficiency at your full load (or anywhere in between).
So in the end it honestly does not matter at this point in time. I’d argue the choice is between price and energy storage (assuming all security considerations are sorted of course.)
Okay I finished vibe-coding a simulation which works under the following assumptions and control strategy.
System
- 11.3 kWp PV system (7.83 kWp East + 3.48 kWp West, or -67 / + 113 degrees to be exact)
- 10.2 kWh BYD battery (LFP)
- 8.17 % roof
- 14% loss
- Bundle includes Fronius Symo Gen24 10.0 plus + Fronius Wattpower Home 11J charging station
Assumed consumption + control strategy used in the simulation
- Base load 450 W during summer, 550W during winter
- Boiler controlled in 1 kW steps, the surplus between the steps goes to the battery. Demand 5 kWh / day.
- Ground source heat pump domestic hot water heating during summer taken from PV + battery if the PV production is such that the daily demand of 1.6 kWh / day can be taken in one 1h period from the battery.
- EV arrives at home at 16:30. Assumed to be home all days weekends. Assumed consumption 20 kWh / day during weekdays, 0 kWh / weekends. Car is only charged to 80% SOC.
- Remaining surplus is exported.
- During winter, battery is charged during cheap night time and discharged during days for base load.
- Battery is used between 15% - 90% SOC i.e. not to extremes so that the degradation / lifecycle would be optimized.
Data used in the simulation
- PV production data with our home coordinates + roof parameters is read from EU PV GIS data for the years 2020 and 2021. The yearly variance is small, so I took the mean of 2 years.
- Avoided import from grid is using the actual spot price / grid tariff that we have paid on 2024-2025 (mean of these). Accuracy is 1h and the battery will smoothen clouds.
Assumptions / speculations on price development
- Electricity prices will go up 2.5% / year (pure speculation, but this is the number I used)
- PV degradation 0.5 % / year
- Battery efficiency losses 1% / year
Simulation results
PV production (kWh)
Jan 45.0
Feb 224.2
Mar 609.7
Apr 1091.3
May 1462.1
Jun 1753.1
Jul 1567.8
Aug 1116.7
Sep 670.6
Oct 248.5
Nov 68.9
Dec 23.0
Total 8880.9
Averages per year (mean of 2 year simulation)
1. PV Production
Total production: 8 881 kWh
Self consumed: 4 371 kWh
Exported: 4 510 kWh
Export revenue: 226.2 €
2. Baseload
Demand: 3 996 kWh
PV: 1 497 kWh
PV via battery: 797 kWh
PV savings: 226.3 €
Battery hedging savings: 43.4 €
Savings: 269.8 €
3. Boiler
Demand: 1 828 kWh
PV: 1 119 kWh
Savings: 113.4 €
4. EV
Demand: 5 220 kWh
PV: 635 kWh
Savings: 46.3 €
5. Battery discretionary headroom (Mar-Sep sunrise)
Headroom: 331 kWh
Rate: 9.9 c/kWh
Savings: 32.8 €
6. GSHP DHW pre-heat (daytime, PV + battery only)
Demand: 323 kWh
PV: 0 kWh
Battery: 323 kWh
Savings: 25.8 €
7. Total savings
Total savings: 714.4 €
Payback
Initial investment: 11 250 €
Year-1 savings baseline: 714.4 € (mean of 2 calendar years at 2024–2025 prices)
Electricity price escalation: 2.5% per year
PV output degradation: 0.5% per year
Battery capacity degradation: 1.0% per year
Projection horizon: 25 years
Simple payback: 14.7 years
IRR (25-year): 4.8%
Year Annual savings (€) Cumulative savings (€) Net benefit (€)
1 714.4 714.4 -10535.6
2 721.3 1435.7 -9814.3
3 728.3 2163.9 -9086.1
4 735.3 2899.2 -8350.8
5 742.4 3641.7 -7608.3
6 749.6 4391.3 -6858.7
7 756.9 5148.1 -6101.9
8 764.2 5912.3 -5337.7
9 771.6 6683.9 -4566.1
10 779.0 7463.0 -3787.0
11 786.6 8249.5 -3000.5
12 794.2 9043.7 -2206.3
13 801.9 9845.6 -1404.4
14 809.6 10655.3 -594.7
15 817.5 11472.7 222.7
16 825.4 12298.1 1048.1
17 833.4 13131.5 1881.5
18 841.4 13972.9 2722.9
19 849.6 14822.5 3572.5
20 857.8 15680.3 4430.3
21 866.1 16546.4 5296.4
22 874.5 17420.9 6170.9
23 882.9 18303.8 7053.8
24 891.5 19195.3 7945.3
25 900.1 20095.4 8845.4
Don’t want to destroy your calculation @masipila, but some additional things should be taken into account to adjust the results:
- 1% annual degradation for the battery is too low. Make it 3% and a full replacement after 20 years.
- Maintenance costs for small installations should be assumed between 3~5% of the initial costs. E.g. assume the inverter needs to be replaced every 12~15 years (which is very realistic). Allow for an increase of this costs also at a rate of 2.5% p.a.
I think, this can also be calculated using a spreadsheet of your choice, that might be more transparent then some vibe coded stuff 