Last summer I became curious about trying hydroponics to grow basic greens on our patio. I looked at potential systems, but the UK pricing for solutions was expensive. For tower systems, these would typically cost £300-£500+ for a tower capable of growing 12-18 plants. This was more than I was prepared to spend to find out if it was something I was interested in.

Earlier in the year I purchased a Bambu X1C that I ran alongside an older Prusa Mk3S. So I thought I would see if I could build something myself.

I’ve recently been returning to writing some software. I was looking for an automated solution to allow me to do the following things :

• Track Vulnerabilities
• Track Licenses Used
• Track Package Versions

I also wanted something that would alert me to new vulnerabilities appearing even when I’ve not performed a build on the project for a while. Ideally, this would be free until I find a revenue stream for the software.

Apologies for the lack of updates on the projects for managing solar, but something happened in January that changed everything.

My original solar installer wrote to us suggesting a battery upgrade for the house. They provided a lot of data on what we had used and when since the installation. They had a payback that was at around three years. Since I’d done a lot of the maths the previous year, I pushed all the data into a spreadsheet and it became obvious that this was the way forward for us.

It’s got to that time of year when the nights are drawing in and the central heating is back on. The shorter gloomier days also mean that the abundance of solar generation we had through the summer is over and my mind turns back to how to be as efficient as possible with what’s created.

Last year I discussed some of the automations I had created through Home Assistant to help reduce energy costs. The basic idea was to try to “turn down” electricity and gas usage in rooms that were not in use. This was achieved through a combination of motion sensors and examining light states. (No motion plus no lights on equals an empty room).

I’m writing this up after wasting over eight hours wrestling with an issue today.

My problem was straightforward. I had created an object that had a void in it. The idea was to pause the print at a specific layer, add a magnet into the created void and then allow the print to continue.

A simplified version is to think about a cube with a smaller cube removed from the centre. The printer prints up to the top level of the central cube and pauses while a magnet is placed in the centre. Once the magnet is in place it continues and completes the large cube.

I can’t justify the extra expense of adding significant additional battery capacity to my solar installation with its current pricing. However, it’s interesting to note that having crunched the numbers, vehicle-to-home would have virtually eliminated all our grid consumption through the summer. It would only have been days when we needed to perform longer trips in the car on repeated days that would have caused us to import any energy at all.

Looking at optimising solar usage in the house is pointless without examining the background power draw of the house. If the background usage is high, we will burn through the energy stored in the house battery and start importing from the grid once the sun goes in.

The stored data/live views of the house can show the overall background power consumption, but not the specific detail. To avoid differences in consumption caused by appliances like TVs and cookers, I looked at the data when everyone was asleep.

After thinking about how to use the solar power my home generates in the most efficient way I realised that this is a generic problem. Effectively we have three types of devices in play :

• Generation Sources. eg. my solar panels, the house battery, the grid.
• Switchable Consuming Devices. eg. My car, a heat pump, the house battery. The load generated by these devices can be turned on/off.
• Background Load. The overall usage at the site. eg. my house load.

This general abstraction is a way of trying to come up with a common interface for each type of device so that we can easily add new devices as households get more connected.

After creating the monitoring system for our solar installation earlier in the year I began to notice a few things about how it performed.

• The solar output of the system is 9.6kWh and our house battery is around 10kWh.
• The battery system has a maximum charge rate that the panels often exceed.
• Whilst we get credit for energy exported back to the grid, the payment credit per kWh exported is approximately 8x less than the charge for 1kWh imported.

So as we rolled into summer, we got into the habit of waiting for peak sun to turn on dishwashers/washing machines and the car charger.

Now that we can pull the data from the local solar inverters, we need somewhere to store it. The excellent solismon3 discussed in the last post presents a simple page that Prometheus can consume to store the historic data.

The reporting partner to Prometheus is Grafana. This allows the creation of graphs/basic reporting of the data like this :

To bundle everything together I created two solismon3 docker images with the configuration for each inverter added. I then used docker-compose to run those along with Prometheus and Grafana as follows :