We’ve published a new blog on smarthomegeeks.io last week, on the energy transition and how individual households can contribute to this. It’s currently in Dutch, as the trigger for this was the announcements of several energy providers in Netherlands who are considering to charge a ‘return-to-grid fee’ for solar panael owners. We provide specific context to the Dutch energy market and draw a couple of conclusions:
It is not possible to go off-grid (yet). For my household, it would require a 3mWh battery to get me through the months November - December (and this battery I call “the grid”).
Grid balance starts locally. During the summer months I could be self-sufficient if I can store the energy my solar panels produce. On the sunny days, my solar panels produce more than my household consumes, yet due to timing I still deliver to and consume from the grid.
A smarthome helps the transition. No need to explain that here
A home battery is unavoidable. As we simply cannot influence time through automation
Unfortunately, financially in Netherlands it is not viable due to legislation that allows to net substract your total delivery to grid over the contract period (saldering).
Now… as we do believe that a home battery is going to contribute, we will start exploring the possibilities of further smart automating our appliance usage and explore calculations to determine the size of battery needed for our home.
As the application of a home battery is a lot more common in Germany and Belgium, we are curious to learn from your experiences and pros/cons in the usage of home batteries and prepare ourselves best for this inevitable move in The Netherlands.
Nice that you’re looking into this. I am myself also working on several aspects of ‘smart’ energy as part of my research and my job. The methods you describe in your article fall in the category ‘rule based control’ or RBC for short. It’s a nice starting point, but far from optimality. To truly get the most out of a smart home you need an energy management system or EMS for short. Such a system uses mathematical optimization to find optimal control curves for each appliance. With this method you can save up to 40% more on your electricity bill compared to RBC. I am myself mostly working on grid-aware heat pump control and solar forecasting, but in the future I will also look at batteries. For now they are simply not economically viable
Thanks for your reply! I’d be super interested to learn more on your experiences with EMS. Completely agree that RBC is the basics, I experience that most of us are still beginning to explore this.
I’ve also included the solar forecast as well as the dynamic energy rates in my HA, and would love to explore the possibilities that HA could bring in the area of EMS.
A ‘real’ EMS doesn’t exist yet. At least not one that ticks all the boxes. There is EMHASS: An Energy Management for Home Assistant, but it is somewhat limited in functionality and technically challenging to setup. Then there are lots of companies selling an EMS as part of their product range, such as solaredge. But the EMS then only works with their batteries and their PV and a selection of external products.
There are upcoming standards to standardize communication between an EMS and an energy consumer / producer, mainly S2 by TNO and EEBUS. Whether one of them will survive, or they will somehow merge, is not sure. IMHO the EU needs to write up a directive to force companies to support such a protocol, because the market is way too fragmented at the moment.
For PV forecasting you can ofcourse use the solar.forecast integration. But I found it lacking in configurability, accuracy and it’s quite an unreliable cloud service (but it’s free, so can you really complain…?) Anyway, I started pvcast for these reasons. With pvcast you can build a model of your PV system and do local solar forecasting with any weather forecast as input.
Great food for thought. We’d like to stay away from vendor lock in as much as possible, so a solaredge solution could be a nice “point solution” or a trial, however will not last in the long run…
Will look into your suggestions for EMHASS and your pvcast suggestion!
