House that is almost energy independent, and is designed for a temperate climate. And some musings on ammonia and nuclear power.

I noticed that photovoltaics are really cheap now. 19.8 kW of peak power would cost around 20 000 PLN (1 USD is equal to 3.65 PLN when I am writing this). Array of 44 of those, with peak power of 475 W each, which actually will be closer to 450 W, because of marketing and degradation progressing with time, will produce aforementioned amount of total peak electrical power. I accounted only for the initial fast degradation, it will slowly get even worse. Appropriate hybrid inverter costs 11 000 PLN.

I was actually thinking about array of 4 rows and 12 columns of 1757 × 1134 mm panels, that are mounted on a south facing side of a gable roof with a 45° pitch. Four places in this array are empty, because of windows that are to be mounted there. Ideally PV panels would track the sun, but in winter sun changes position in the sky only slightly, and pointing photovoltaics directly at it is not a bright idea, because much of the light received is diffused, and comes from around the sun, and rather not from below the horizon.

30 kWh of LiFePO4 batteries would cost around 14 000 PLN. This would be enough to provide 8 kWh of electricity during most of the days, even during Polish winters. 12 kWh of electricity would be much trickier in off-grid scenario during November, December, and January. I am assuming that only 60% of maximum batteries capacity is used (overcharging and deep discharges are bad for batteries). I would like to add that if all of the lithium produced during a year was to be converted into LiFePO4 batteries, and then all of them would be distributed equally among all humans, everyone would get only around 0.3 kWh of them. So per a member of 4-person household 25 years of lithium production would be required.

The real problem however is heating of home during winter. My idea for solving this is using electricity that cannot be stored in batteries during sunny winter weather to power cheap air source heat pumps to increase temperature of a "swimming pool", that is actually a thermal energy reservoir consisting of 50 tons of water. Reservoir is slightly insulated from the rest of the building. Three heat pumps with COP of 4.15 and nominal heating power of 16 kW at A7/W35 costing 9000 PLN each would suffice to fully take advantage of winter sun. They would have COP of 3.25 and 15 kW of capacity in the A7/W45 scenario. Number after the letter “A” is temperature of outside air, number after letter “W” is temperature of heated water, in degrees Celsius. I am pretty sure that outside air temperature during sunny wether will be above freezing (I used “HOURLY DATA” part of this tool to check that). Temperature in the “swimming pool” during heating season would vary between 30 and 40 degrees Celsius. It is equal to the total heat capacity of the reservoir of 580 kWh. If “swimming pool” is located in a rectangular room with dimensions of 6 m, 5 m, and 3 m, then when fully charged (40 °C), and insulated from the rest of the building, that is held at constant temperature of 20 °C, by the styrofoam with a thickness of 8 cm, and thermal conductivity of 0.035 W/(m⋅K), it would self discharge at a rate of 1.1 kW.

In the case of a southern Poland, during an average December day, 12.3 kWh of power would not be stored or directly consumed as electricity, in the scenario of 8 kWh electricity per day consumption. This could be used to make approximately 40 kWh of heat. Well insulated building with recuperators, and possibly ground-coupled heat exchanger for the forced ventilation’s incoming air (located below house’s foundation insulation), could probably keep a semi-decedent temperature inside with that kind of heating, although keeping everyone comfortable would require additional energy from the outside.

House in which I currently live uses twice as much heat in winter and is smaller than the construction that I am currently designing. But new construction will have thicker layer of insulation and more energy efficient ventilation. “Swimming pool” will self discharge completely over a period of more than a month, it should be possible to fully charge it during sunnier months preceding harsh winter, reducing energy requirement during that time. In November 20.0 kWh per day would not be captured by “intended” electrical installation and will be available for heat pumps. In October 37.4 kWh per day should be expected.

I need to add that 8 kWh of electricity per day is not that much, especially for a construction with forced ventilation, but if amount of computations done and time spent in front of large and bright screens is reduced, then it should be survivable.

When it comes to producing domestic hot water, I would like to use waste heat from water-cooled energy demanding electronics, such as CPUs and GPUs, to generate heat required. Refrigerator (which basically is a heat pump) could lower temperature of electronic components, while increasing temperature of utility water. Additional source of heating might sometimes be required (extra heat pump for instance). If tank with utility hot water gets too hot during winter, it could transfer heat to the "swimming pool".

Large heat pumps and "swimming pool" could be used to keep interior of the house cool during summer. Amount of electricity produced by photovoltaics and not consumed by regular appliances will be equal to the 60.6 kWh per day in July, even when regular appliances start to consume 12 kWh per day. Electricity can also be sold during overproduction period, but price will probably not be that high. Electrolysis of hydrogen is also an option, but a problematic one, more on this later.

“Swimming pool” could potentially be used for swimming. But temperature may not be comfortable, access will be through a small hatch, and keeping the water clean will be much more challenging.

I would like to add, that I think that breeder reactors are better energy sources. They are not intermittent, so production of hydrogen and ammonia would be much cheaper. I dream of a transportation that is ammonia based. 1 trillion (10¹²) PLN would be required to build installations producing 1 EJ (10^18 joules) of chemical energy that can be released during combustion of ammonia over all year, which is the amount required to power all road vehicles in Poland. Approximately 66.3 GW of electricity would be required to keep this operation continuously ruining. So if 1 GW of power from nuclear power could be kept below 7.9 billion (7.9·10⁹) PLN, 1 kWh of ammonia would cost as much as 1 kWh of crude oil. Intermittent solar power would increase cost of hydrogen and ammonia production equipment at lest twice, while being able to produce required electricity at trice the cost. I am assuming that all installations are able to survive for 20 years, and solar is just scaled up version of electricity-generating part of described house, producing 21 MWh per year (previously mentioned tool predicts exactly 21.7 MWh) for 50 000 PLN (mounting and wiring cost some money too, I am probably underestimating how much, so thrice is rounding up of 2.64). Another assumption is that efficiency of electricity to ammonia conversion is 50%.

Nuclear reactors are usually more costly, but they should not be, and sometimes they were not. New car with a 100 kW maximum power of internal combustion engine costs now approximately 100 000 PLN. Scaling it up means that a car with a 1 GW engine would cost 1 billion (10⁹) PLN. Cars are not used all the time, and rarely at maximum power, so multiplying this price by a factor of 7.9 should give a price of an equivalent nuclear reactor made from efficiently mass produced parts.

I will probably add illustrations and additional information in the future.

2026-05-25 EDIT: I did a major overhaul of this text.