Power generation is the business of producing usable energy where you need it, whether as electricity, as shaft power turning a machine, or as heat.
The first question is never how to generate it. It's how much you actually need, and the honest answer is usually far less than the installation being quoted for.
Almost everyone starts at the wrong end
The usual sequence is to decide on a technology, size it from a brochure, install it, and then find out what the household actually uses. That produces systems that are simultaneously too large in capacity and too small in storage, and it's why off-grid power has a reputation for costing more than it should.
The right order reverses that. Measure what you use, cut what you can, work out what genuinely has to run and when, and only then ask what should generate it. Reducing demand is almost always cheaper per unit than generating more, and it keeps being cheaper.
There's also a distinction worth making early: a great deal of what a household calls power is actually heat, and heat is far cheaper made directly than made from electricity. Separating those two changes the size of the problem considerably.
01
Working out what you actually need
This is the step that decides the cost of everything after it, and it takes a week instead of an afternoon.
List every load, what it draws, and how long it runs. Multiply and add, and you have daily consumption. Then look at when those loads happen, because a system is sized by the coincidence of demand as much as by the total, and a household that runs its heavy loads at different times needs a smaller system than one that runs them together.
Then separate the loads into three groups. What must run continuously, refrigeration, some medical equipment, water pumping. What matters but can be scheduled, laundry, tools, charging. And what's convenience, which is where reductions cost nothing.
Efficiency first, always. Replacing an inefficient load is nearly always cheaper than generating the power it wastes, and the saving repeats forever. Lighting, refrigeration, pumping and standby loads are the usual places the waste is.
And be honest about the worst case and not the average. A system sized on average demand fails in the week you most need it.
02
Mechanical power, which skips a conversion
Not everything needs electricity, and turning a shaft directly is more efficient than generating power and then using it to turn a shaft.
A great deal of historical machinery was driven by belts and line shafting from a single power source: mills, pumps, saws, grinders, workshops. That arrangement is still sound, and it removes generators, controllers, batteries and inverters from the chain entirely.
Water pumping is the clearest case. A wind pump or a water-powered ram moves water with no electricity anywhere in the system, no storage and almost no maintenance, and both were ordinary farm equipment for a century.
The trade is flexibility. Mechanical power is available where the shaft is and when the source is running, where electricity can be moved and stored. For a fixed load near a steady source, mechanical wins clearly.
03
Engines and the fuels that run them
Engine generators are the fallback almost every off-grid system keeps, because they produce power on demand regardless of weather.
Diesel engines are efficient, long-lived, and tolerant of being worked hard, and they dislike running lightly loaded for long periods. Petrol engines are cheaper and lighter and generally shorter-lived. Both need fuel, which is the dependency.
Fuel storage is its own problem. Petrol degrades within months, diesel keeps longer but grows biological contamination in the presence of water, and both need proper containers, ventilation and separation from ignition sources.
Alternative fuels change the dependency in place of removing it. Wood gas, produced by heating wood with restricted air, ran vehicles at national scale during fuel shortages and works; it's laborious and produces carbon monoxide, which makes it an outdoor process with real hazards. Ethanol and vegetable oil derivatives both work with engine adaptation.
Run a generator properly loaded and infrequently and not continuously and lightly. That's the pattern that suits both the machine and the fuel budget.
04
Storage, and why it dominates the budget
Generation is the easy part. Having power at three in the morning after four grey still days is the expensive part.
Battery capacity is not the number to compare. What matters is cycle life, how deeply the bank can be drawn without damage, and how it behaves in your temperature range, because those together determine what each stored unit actually costs across the life of the bank.
Non-electrical storage is frequently better value where it fits. Heat in water or masonry, water pumped uphill during a surplus, cold stored as ice, and fuel itself are all storage, and none needs a battery.
The cheapest storage of all is not needing it. Moving loads to when generation is happening costs nothing and removes them from the storage calculation entirely, and a household that does its heavy work in the sun or the wind needs a fraction of the bank of one that doesn't.
05
Getting it to where it is used
Distribution is where efficiency quietly leaks away and where the safety requirements are non-negotiable.
Voltage and distance interact: the lower the voltage, the higher the current for the same power, and the more is lost as heat in the cable. Low-voltage systems need much heavier cable over any distance, and undersized cable is both a loss and a fire risk.
Conversion costs something at every step. Every inverter, charger and controller takes a percentage, and a system with several conversions in series delivers noticeably less than it generates. Arranging loads to use power in the form it's generated avoids that.
Standby losses are the ones nobody accounts for. Inverters, controllers and always-on equipment draw continuously, and over a day that can rival a real load.
And the fixed wiring is a regulated, life-safety matter. Generating your own power changes nothing about that, and this is a place where the right answer is a qualified person.
Distribution, redundancy and what happens when one source fails, which is what makes a set of machines a system.
Where this subject stands
The library doesn't cover Power Generation yet. It's in planning, and what gets written next is impacted by what readers ask for.
Tell us what you actually need to run and what sources you have, sun, wind, water, fuel or a mix. Sizing follows from the loads, not the other way round.
Readers asking for this
0marks
No details needed. Counted on this device: the shared tally is not live yet.
Questions
How big a system do I need?
Smaller than most quotes assume, if you do the demand work first. Measure what you actually use, cut what you can, separate loads into must-run, schedulable and convenience, then size to the worst case and not the average. Efficiency is nearly always cheaper per unit than generation, and the saving repeats forever.
Do I need batteries?
Only for power at times when nothing's generating. A great deal of household activity can be moved to when the sun or the wind is available, and every load moved is storage you don't buy. Where storage is genuinely needed it doesn't have to be electrical: heat, cold and lifted water all store energy without a battery in the chain.
Can I connect a generator to my house wiring?
Only through a proper transfer arrangement that makes it physically impossible to feed power back into the incoming supply. Backfeeding can kill someone working on the line, and it's both a safety matter and a legal one in most places. This is the part of the subject where doing it yourself is the wrong answer.
What is the most reliable off-grid power source?
Hydro, where you have falling water, by a wide margin: it runs at night, in the dark and through still weather, which means it needs a fraction of the storage the others do. Failing that, the most reliable arrangement is usually more than one source with different failure modes, plus an engine as the fallback that doesn't care about the weather.
The Ark is a library of practical knowledge that was ordinary once and isn't any more, being recovered from the sources that recorded it while it was still in use. Membership is what pays for that work, and it's what decides how fast subjects like this one get written.