Solar panel to DC heater
In Norway electricity is the most popular form of heating for households at 96% (source). Pure 230V AC is a very high-level energy, and we don’t really need that precision for heating.
This is a project to see if easy, not-perfect, locally generated energy can be used to lower the heating bill.
Disclamer: this is mostly for fun. A more economic aproach would be insulation, heat pump, put on a sweater, etc.
The plan
┌──────────────────┐
│ │
│ │
┌─────┐ │ ┌──────────┐ │
│ │ │ │ │ │
│ PV ├───┼──┤ Heater │ │
│ │ │ │ │ │
└─────┘ │ └──────────┘ │
└──────────────────┘
House
A solar panel, with no controller or inverter, and a heater. Dumping heat in a basement that needs heating all year round.
The problem
Solar panels have an optimal voltage, where they produce most power, and this changes based on light conditions. A proper controller (MPPT) is expensive, so I want to see if I can do without.
Hardware
Specs of my first setup:
- 180 Watt panel (the most power I could get in a normal 12 Volt range)
- 100 Watt heater with fan (I’ve ordered a 180 W model too)
- ESP8266 for statistics (with dht22, ads1115, acs712)
Setup
The panel, wall mounted in a very sub-optimal location (shadows from people, walls).
- Panel connections in
- Current sensor
- Relay to switch off
- Connections out to load
The heater with fan. It’s tiny. The heater is a PTC, so I don’t think it will pose a threat if the fan dies. In front of it is a white temperature sensor, to see that we actually produce heat.
A micro controller, with an external ADC. The micro controller generates graphs and control the relay. The resistors by the ADC are a voltage divider, to scale down the ~23 Volts the panel can deliver to a safe 5 Volts the ADC can measure.
Unloaded the panel will be at ~22 V. With this load one can see that the output is only good when the potential keeps between 12-14 V. I have no way of controlling the voltage, except for changing to a smaller load (less effective heater).
Graph of power output. Totally this gave around 600 Wh on a mostly sunny day.
Green line is the temperature sensor next to the heater. Measuring right next to the fan, I’ve seen it at ~60 °C, but this one is placed a bit away. The orange line is a temperature sensor elsewhere in the room. The room seems to rise about 2 °C during a day. The gray field is cloudyness in percent.
Conclution
So far this seems to generate heat as it should, but it will have to run for a pretty long time to earn back the ~1700 NOK I paid for the panel. I will test with different heaters to see if I find something that works better. A posibility is also to have several heater that can be switched on with relays based on the power available. A kind of very crude MPPT.
Thermal
The panel seems to have some significant hotspots. Normal? Cheap panel?
The relay is producing heat, but waste heat isn’t a problem in this case. =)
The relay is pulling more current than a GPIO from the ESP is rated for, so it’s voltage regulator is getting very hot. Should probably fix that.