Results after 1 year of simulation

[cybermaus]

I have run the simulation for a year. a 20kWh battery on a 24kWp solar installation and a annual yearly use of 6MWh
Most of that goes into heating using the heat pump

What you can see, and what I had expected, is that the battery is mostly useful in spring and autumn. the use curves go flat. In summer, the battery is soo full, and the night draining it so short, most capacity is unused. In winter, the battery is soo empty, and the days filling it so short, most capacity is unused.

All in all it means over the course of a year, I merely save 1500kWh. Mostly in spring and autumn.
We really need something to lift the solar energy from summer to winter, and not merely from today to tomorrow.

Here is the graph with SoC of the battery. Note that mid-Januari I changed it from 10 to 20 kWh Also note that in summer, the battery capacity is pretty much not relevant as it is always full. Winter is a little less clear, but it is mostly empty, but there are a few more occasional good days.

The transition where it hits both full and empty is where the real work is done. For example, doubling the battery in January didn't have any effect till March.

[hif2k1]

It's interesting isn't it. What we really need is seasonal energy storage from summer to winter! If only that was possible.

Did you try different sized batteries? It seems there are diminishing returns on a bigger battery. A 5kwh battery may save you almost as much energy as a 20kwh. The only advantage in a big battery is if you are wanting to charge from the grid at cheaper/greener times of day.

[dewi-ny-je]

Getting co2 from the atmosphere is a very energy intensive process, keep in mind that co2 is about 450 ppm, not 30% volume.
It means the efficiency is very low, a lot of heat is wasted on the atmosphere.
Liquid fuels are essential for air transportation, but that's about it.

Also such a complex plant cannot run 7 hours a day in summer, or whenever there is surplus of wind. It has to run 24/7, which means it can't run on renewables.

[cybermaus]

Yes, they are not easy to make, especially the larger molecules, hence focussing on alcohol, for which we have a cheat in the form of fermenting plant matter. We could also do oil from rapeseed of course, but then we get into the debate about food supply.
In any case, it is cyclic CO2, which was the point countered.

Going back an earlier point, I still feel we should go for low-risk low-tech solutions like pumped gravity storage. I am not entirely sure what the latest tech is on hydrogen, but I still worry about its pressures, cooling, and risk of explosions, especially when we have caches of it large enough to make a dent into the energy budget.

[kluner]

If it is synthetic liquid fuel, we are not back to pumping CO2, we are recycling CO2, with a net zero effect.

Liquid fuels are not the problem, fossil liquid fuel is.

You're forgetting the other problems with burning liquid fuels: they generate pollutants. NOx primarily.

There will always be applications where liquid fuels are acceptable despite that facfor. But they will be the exception. For most things, electricity will be the far better choice.

[rrozema]

It's interesting isn't it. What we really need is seasonal energy storage from summer to winter! If only that was possible.

Did you try different sized batteries? It seems there are diminishing returns on a bigger battery. A 5kwh battery may save you almost as much energy as a 20kwh. The only advantage in a big battery is if you are wanting to charge from the grid at cheaper/greener times of day.

The fact that returns diminish on a bigger battery is easily explained if you think about it: every battery has a limit to its storage capacity and once your daily production exceeds you battery's capacity, you can only charge as much as you've emptied the battery the night before. If the battery is smaller than the amount you use during the night (for example during winter time), you will run the battery empty and the next day you can fill it up to 100% (or whatever is available). But if the battery's capacity is over what you're using during the night, you will not empty the battery and thus you won't be able to store any more than you discharged the night before. Unless you're doing a 'sale' during the night (= force discharge into the grid, preferably when the export price is higher than the costs taken for charging during the day) you will ever only be able to charge as much as you've used during the night and thus returns will top off at this point, no matter how big a battery you put in.

That knowledge of matching the maximum capacity of the battery with the nightly usage was my first observation from the simulations I ran. Another observation was the capacity of the inverter: smaller batteries tend to be able to deliver less power (W). If a load takes more than the inverter can output, you're still importing energy from the grid, even if the battery is not empty. This is a balancing act however: if you want to be able to support all of your devices' loads you need a big and expensive battery, but you're only going to use that higher output a very small percentage of the time: it is often cheaper to simply import the remainder (that which the inverter can't deliver) from the grid. The next observation was that it matters WHEN you charge the battery: if you leave the battery in default mode, it will start charging as soon as the sun rises until the battery is full. At this time however prices for exported energy are still relatively high, whereas around noon the prices are usually at their lowest. For this reason I created a sensor that keeps track of the charging costs: the price in EUR/kWh for the energy in the battery. By postponing the charging of the battery to the time when the export price is lowest I managed to lower that cost of charging the battery and thereby significantly increase the return from discharging the battery when the price for importing energy is high. The same battery capacity, the same amount of energy saved, but higher returns.

I now know that I need for my home and my family's routines roughly 8kWh capacity, and that I need -during summer- only 2 to 3 hours to charge that amount of energy. I want to further extend my experiments to predicting how much time I will need to charge the battery based on season and/or weather forecast, so I can automate finding the optimal time to charge the battery. Plus I have the option to configure my solar panels (Enphase) to reduce production so that nothing gets exported when production exceeds my usage. In summer my solar capacity exceeds my daily usage by 3-4 times, during winter however many days it comes close to my daily usage. By reducing the output during summer I avoid sending excessive amounts of energy into the grid, whereas during winter I still get reasonable returns.

[dewi-ny-je]

I would have said it in a different way: a small battery, which is in summer basically always fully charged with excess production, and which is completely depleted at night with the bare self consumption (in summer!) allows whole of its energy to be used to avoid purchase, and therefore the value of the energy stored it the energy purchase price, which is basically always higher than the value of energy sold to the network. Also, as stated, smaller batteries have lower output power, so that they run at higher load% during discharge and the efficiency is higher (pull 200 W from a battery able to peak at 4000 W and you are loading it at 5%... the efficiency is never great).

If a larger battery is purchased, the extra energy won't be always available from the solar panels, and also this extra energy cannot be used at night, but it can be sold to the network, with a smaller economical return, so the ratio "economical revenue/battery capacity" will be overall lower.
If the battery gets even larger, a larger number of days/year the battery will not get filled up making the economical return even lower.

Now, this is the economical return which is always increasing less than the battery capacity. However, the purchase price also increases less than the capacity: twice as big is rarely twice as expensive (exept for Marstek I think).
The ratio purchase price/economical revenue cannot be easily defined, unless someone runs a simulation (in parallel) of two or more batteries of different capacities to track revenue with capacity.

[cybermaus]

Yes, the graph is in fact 10kWh for the first 9 months, and 20kWh for the last 3 months. Adding yet more will indeed add more investment, but only add a week or so to each of the 4 transitions from spring/authumn into summer/winter. I think the ideal value is to be able to overcome 2 bad sun days, meaning 2 day usages.

Biggest problem is estimating how much it means in finances. The investment is easily calculated but the returns less so, especially because the energy companies obfuscate and fluctuate (over the years) the way they calculate.

I think purely financially, it is a bad investment for a typical winter-heating summer-nothing household.
But technically, as a nerd, it is hard to resist.

[dewi-ny-je]

Thanks for the feedback!
I'm on mobile so I won't give you my data now, but in the meantime can you explain what we're the rules for charging and discharging?

Also, was it with current tariffs or, if you are from the NL, with the ones after subsidies will be stopped?

[cybermaus]

I did not set up any rules, as I did not attempt to trade energy. Just use my own surplus and try to prevent energy import/export.

So rules were the default as installed: Solar energy surplus means charge, and solar energy deficit means discharge.

I did set 3kW per phase max charge/discharge current, totalling 9kW in total, as I imagined the 3kW Victron Multiplus model. I did lower the expected efficiency from 95% to 85%, dropped the first 5% because I do not believe the 95% vendors advertise with, and the second 5% because with Victron I would be going for low voltage batteries.

[dewi-ny-je]

I have been simulating a battery since beginning of March.
I assumed Dutch rules after subsidies will be revoked, January 2027, since I was thinking about a battery next year (but in fact I just pre-ordered a Zenduee Solarflow 3000 mix).

I have 4 kW solar (maximum recorded production power) and about 3000 kWh imports, 2500 kWh exports, 3700 kWh solar generation.

I simulated a Marstek Venus 5 kWh, and I also introduced in it efficiency curves for charge and discharge based on the values measured by a couple of people online, so that slow charge or slow discharge correctly result in more losses (the feature will be in battery_sim as soon as I fire my laptop once I'm back home from holidays).

Oh, I have a dynamic contract, which means I pay/get paid for electricity the bare energy price on the spot market plus 2 cents per kWh (it gets paid back when I export).
This contract for my household is the cheapest contract currently available, even if solar energy I'm currently exporting is worth little to nothing because... everyone exports at the same time (what a wonderful energy source...).
It also means it's very good for batteries because the missing income is very low and the value for the energy discharged high, evening and morning peaks.

In 2027 purchase involves paying taxes, exporting means not getting them back, so the gains from a battery are pretty high ASSUMING I'M SELF CONSUMING IT, while energy stored at midday and expected to the network in the evening only benefit from the difference spot prices. In other words: a battery larger than my self consumption at night has much reduced benefits. For me it's 3-4 kWh.

My results are that so far I had an economical return of 1 euro a day, which is pretty amazing.
Days with flat energy prices bring low gains, 50 cents a day of less, but then in the weekend energy at midday is worth nothing or I even get paid for using it (in that case I turn off my panels, last year I also turned on 4 KW of space heaters in my garden to earn more...) and on the same day the peak prices skyrocket so in a day I can earn even 3 euro. I mean, when that happens one day, it's like raising the average of the week by almost 50 cents a day!

But I didn't use the Default mode (zero flow at the meter, nul op de meter in Dutch), I have some more complex automations.

I record and store in a helper the morning 6-9 consumption (typical morning price peak), and in the evening I limit the automatic discharge to ensure I have energy to power the home, become delivering/discharging at night when energy is cheap and then importing in the morning is stupid. I work on the assumption that yesterday's consumption will be replicated today, which results in very small deviations from the actual value.
At 9 (after morning peak) I check the forecasted solar production during the day. Above a threshold, battery charge is limited to 15% (to have something for unexpected leaks like a tea coffee or something small), and I bring the max SOC to 100 only later (typically 12:00, manually adjustable) so that my panels between 8-12 can export energy, when it's really valuable! (always install panels east/west, not south, to have more production when it's worth something and less at midday when there's not point anyway).
If the forecast is low, the battery is free to charge immediately, better fill it rather than having to import.

In the evening it goes in Default mode until the reserve for the morning peak is reached, which is basically always after the end of the evening peak price.

Now that I'm on holiday and the night consumption is low and the battery would not discharge completely by 9:00, I introduced a new rule which I will improve when back: at the evening peak price I export all the extra energy I don't expect to use (now it's a fixed value, but I'll modify it to use yesterday's value), so that I can have extra earning in the evening: battery cycles are not an issue, I prefer to increase the cycles and the revenue within 10 years.

I think I have another couple of rules but they are minor.

I plan to install EMHASS sometimes, to automate this machine learning of my habits.

The battery I ordered is 8 kWh, so the rules for forcing exports in the evening are important.

About winter: I don't have a large heat pump but I think the daily gains won't go beyond 50 cents. Overall, I expect decent returns over the year.

Not a good investment but far from losing money

[hif2k1]

I agree with you @cybermaus it needs to be simple. I think the most realistic option is a low grade heat store - a large insulated water tank underground which you heat up little by little through the summer and then use directly for heating or with a little help from a heat pump in the winter.

…

On Mon, 11 May 2026 at 06:43, Arjan Koole ***@***.***> wrote: If it is synthetic liquid fuel, we are not back to pumping CO2, we are recycling CO2, with a net zero effect. Liquid fuels are not the problem, fossil liquid fuel is. You're forgetting the other problems with burning liquid fuels: they generate pollutants. NOx primarily. There will always be applications where liquid fuels are acceptable despite that facfor. But they will be the exception. For most things, electricity will be the far better choice. — Reply to this email directly, view it on GitHub <#214 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AFAIZGPFNVQIVU4VL6LJEHD42FR73AVCNFSM6AAAAACYW74SISVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTMOBXGUYTIOA> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you commented.Message ID: ***@***.***>

[dewi-ny-je]

Someone wrote a paper about that, but I don't remember the title. It was about a cylindrical tank to be built next to a house about 3 m diameter and 3 stories high. I think it could work to some degree but I don't remember the details.

I doubt it would be easily repaid in it's lifespan though.

[dewi-ny-je]

It was this one
https://protonsforbreakfast.wordpress.com/2022/07/21/a-sand-battery-not-obviously-a-great-idea/

[cybermaus]

Some Netherlands university linked startup was trying to include some salt/brine phase change into their water based 'battery'.

[hif2k1]

Phase change would help, but even with just water that blog post suggests it could last 6 months. 3.8m by 7m is obviously a big tank, but it's comparable to home swimming pools which are feasible if expensive water tanks. Puting it under ground would help with insulation and make it less ugly.

…

On Mon, 18 May 2026, 05:18 cybermaus, ***@***.***> wrote: Some Netherlands university linked startup was trying to include some salt/brine phase change into their water based 'battery'. — Reply to this email directly, view it on GitHub <#214 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AFAIZGJTHDH6ZUA5AESGQFL43KFI3AVCNFSM6AAAAACYW74SISVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTMOJVGU4TCMY> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you commented.Message ID: ***@***.***>

[dewi-ny-je]

It would require a lot of materials and costs for installing it. Multiply it by all the households and you see the material usage would be huge, and the CO2 emissions required to manufacture those materials as well.
And the houses would get even more expensive, which is already an issue basically everywhere (surely in NL).

Once again, it's adding more costs and emissions to make solar viable, which is already the source with most emissions among the green triad.

We should (as IPCC states in their reports) go for nuclear for the base consumption and limit solar to daily peaks... abandon the expensive and likely hopeless idea of using solar for everything, which is not a reasoned plan, it's just religion/fundamentalism, where you start with the conclusion no matter what, and you find a way to get there.