What a schedule actually is.

A blackout schedule is not a blackout. It is a timetable. Households are put into groups, and each group is told which hours it will have electricity and which hours it will not. When the grid is holding, the timetable is followed and you can plan around it — cook when the power is on, charge everything, fill containers with water. When the grid is hit, the timetable stops meaning anything and the outage becomes an emergency one, which is the kind you cannot plan around at all.

In December 2025 the UN’s human rights monitoring mission in Ukraine reported that most regions were facing scheduled rolling cuts of up to sixteen hours a day, and that after one series of strikes, emergency outages in some areas ran past thirty-six hours. This summer has been easier: in June, the head of Ukraine’s grid operator said the country could expect emergency cuts of up to about five hours a day through July and August. Both of those are real numbers from named sources, and neither of them is a forecast for the coming winter, which nobody can give honestly.

Each of the three figures with its source and its date attached, because presenting the winter numbers as current would be the easiest mistake in the whole piece.

What a small turbine will not do.

This section comes before the other one deliberately. A roof-mounted turbine is a few hundred watts to a few kilowatts in good wind. That is not a generator, and treating it as one gets people hurt.

It will not heat a home. Electric heating is measured in kilowatts held for hours, and no small wind turbine on a domestic roof will carry that in winter. It will not run an electric cooker, an immersion heater or a washing machine at will. It is not a substitute for a hospital, clinic or shelter generator, and anyone specifying power for those should be talking to a diesel supplier, not to us. It also cannot be relied on to produce on any particular day: wind is not a schedule either, and there are calm weeks in winter.

And there is a practical limit that has nothing to do with electricity. Most people living under a schedule live in apartment blocks, where the roof is not theirs, the mounting is not theirs to decide, and a mast on a residential building is a structural and legal question before it is an energy one. This is realistically a thing for houses, not for flats.

A second source does not fix a grid. It changes what a household can still do while the grid is down.

What it will do.

The useful shape is not turbine-to-appliance. It is turbine-to-battery, and battery-to-the-few-things-that-matter. A small wind turbine trickling into a battery bank over a long windy night is doing something a solar panel cannot do at that hour: it is refilling the store while everyone is asleep and the grid is off.

What that store then runs is the modest list: lights, phones and a router, a water pump, a fridge cycling on and off, a laptop, and the charger for whatever else the household depends on. None of that is impressive on a specification sheet. All of it is the difference between an evening that works and an evening that does not.

The wiring is the whole argument. The same turbine is a disappointment on the top row and quietly useful on the bottom one.

Why wind, specifically, in winter.

This is the one argument that genuinely transfers, and it is worth stating precisely rather than generously. Solar has a season; wind mostly does not. In a typical year at Kyiv the December sun delivers about eight per cent of what the July sun delivers, while the wind’s own quietest month is July and even that carries about a third of the energy of its windiest. Wind is not at its best in midwinter — March is windier — but it does not collapse in midwinter either, and sunlight does. A household running on one source has a bad month. A household running on two sources has a bad week, because the two do not fail at the same time.

Checked rather than asserted. February is a genuinely calm month at Kyiv and the chart shows that too, which is the point of drawing it from data instead of from the argument we wanted to make.

That is the same point made in every other post on this blog, and it is not a Ukraine argument at all. It just matters more where the grid is being attacked than where it is merely expensive.

What we are not going to do.

We are not going to publish a customer’s story. We have orders from Ukraine, and every one of them comes with a name, a street and a town attached. Publishing that a particular address has an independent power source is not a testimonial, it is targeting information, and no amount of goodwill makes it safe. Nobody will be named here, ever.

We are not going to run a countdown, a war-themed discount or an urgency campaign. And we are not going to tell you a turbine keeps anybody safe. It runs a light and a router for a few hours. That is worth having and it is not worth exaggerating.

If you are in Ukraine and want to work out whether any of this applies to your building, the useful thing to send us is your situation — house or flat, roof or ground, what you actually need to keep running — and we will tell you honestly if the answer is no. It often is.

The honest version of this is small and dull: a second source, a battery, and a shorter list of things that stop working.
Anyone selling you more than that, on this subject, is selling you something else.

The sixteen-hour and thirty-six-hour figures are from the UN Human Rights Monitoring Mission in Ukraine, published 12 December 2025, and describe last winter rather than the present day. The five-hour figure for July and August 2026 is from Vitaliy Zaichenko, chief executive of Ukrenergo, in an interview with RBC-Ukraine published 25 June 2026. We have given no forecast for the coming winter because we have no basis for one. No customer is named here, no order is described, and no location of any installation is published. The map at the top shows Ukraine’s internationally recognised borders, Crimea included; the public-domain Natural Earth basemap we drew it from does not, so we corrected it, and we say so rather than leave a border quietly redrawn either way. It shows no front line, because we have no source for one worth putting on a page. The monthly sun-and-wind chart is built from the European Commission Joint Research Centre’s PVGIS v5.2 typical meteorological year for Kyiv, with radiation from PVGIS-SARAH2 and wind from ERA5, drawn from 2005 to 2020; wind is plotted as the mean cube of the ten-metre wind speed, which is the energy available in the wind and not the output of any particular turbine, and a typical year composes each month from a different real year. Turbine output depends entirely on the wind at your own site and the figures on our product pages are our own published data.

What you are looking at, in order.

The first shot is a screen, not a machine, and that is where every part actually starts. The red and yellow shapes are the parts, laid out on a sheet. Arranging them so that as little steel as possible is left over is called nesting, and it is the least glamorous cost saving in manufacturing: the same parts, the same machine, the same minutes, and a smaller bill for material because somebody spent ten minutes packing the drawing tighter.

Then the bed. The black ridged strips the sheet rests on are slats, and they are consumable — the beam cuts into them a little every time, and eventually they get replaced. The sheet sits on top and does not move. The head moves.

Then the part that is worth watching twice: the sheet after the cut, with the parts still sitting in their own outlines, and the slats underneath glowing orange. That glow is the heat that went into the steel and did not leave with the part. It is also the reason the shot exists — a cut edge photographs as an ordinary line five minutes later, and looks like this for about thirty seconds.

The next bench along. Parts cut on that machine end up inside these.

The boring part is the whole point.

Nothing in that clip is dramatic, and that is the argument. When somebody else cuts your parts you pay for four things: the part, their margin, the freight, and the wait. You also accept their minimum order, which is how you end up buying five hundred of something because that is the smallest run they will start, and carrying the other four hundred and twenty on a shelf until you need them.

Cutting them here removes the margin, removes the freight, removes the minimum order, and shortens the wait from weeks to that afternoon. It is not a clever strategy, it is arithmetic, and it is most of the reason a home wind turbine from us costs what it does. The longer version of that argument, including the parts we still buy from people who make them better than we would, is in what buying a machine actually buys you.

A machine that runs is an asset. A machine that runs one day a week is a more expensive way of buying the part.

What the video does not show you.

It does not show the invoice, the maintenance contract, the extraction filters, the compressed air, or the operator who had to be trained before any of it produced a usable part. It does not show the days the machine sits still. A laser only beats outsourcing at volume, and the honest version of this post says so: we buy machines when the work is already there, not to have something impressive to film.

It also does not show quality control, which is the half of the job that happens after the cutting. A part being the right shape on the day it was made proves very little. Somebody has to keep proving it, on their own equipment, on a routine, at a cost nobody enjoys paying.

The short version.

Same footage, cut for a phone, if that is where you would rather watch it.

In a year that machine will be an unremarkable thing in the corner of a workshop, covered in swarf, with somebody’s coffee on the control cabinet.
The only trace of it will be a price that did not go up.

The footage is our own, filmed at our own factory on an ordinary production run, and is unedited apart from trimming, stabilisation and colour correction. The machine is a BODOR fibre laser that has been in service for about two years. No customer is named here and no order details are published. We have given no figure for what the machine cost or what it saves, because we could not show you the calculation — and a number you cannot check is worth nothing. Turbine specifications referenced are our own published data and are checkable on the product page.

That last point is usually something a manufacturer would bury. We are going to lead with it, because for a classroom it is not a shortcoming. It is the entire reason these are worth having.

What is actually in the box.

Each crate holds a complete set: the turbine itself, its blade set, and a charge controller with a dump load. Nothing has been removed and nothing has been used. They have simply sat in a dry warehouse since 2023 while we shipped newer stock.

The dump load is worth a paragraph of its own, because it is the part people do not expect and the part a physics teacher will like most. A wind turbine that is spinning has to push its energy somewhere. If the battery is full and the machine is left connected to nothing, there is no electrical resistance to slow the rotor and it will run away with itself. The dump load is a resistor that takes the surplus and turns it into heat. It is a safety device, and it is also the most visible piece of physics in the whole system: when the batteries are full, the dump load gets warm, and a class can put a hand near it and understand conservation of energy without anyone drawing a diagram.

Forty complete sets. The third number is the one that matters for a school.

The thing that dates them is the thing that suits a classroom.

These are not connected devices. There is no radio, no app, no cloud account and no dashboard. In 2023 that was ordinary. Today it reads as old-fashioned. For a school it is three separate advantages, and none of them are marketing.

The first is paperwork you do not have to do. A connected device on school property means somebody has to answer what it transmits, where that data goes, and who processes it. A machine with no radio raises none of those questions, because there is nothing to ask about.

The second is that nothing can be taken away from you. Connected hardware stops working when the manufacturer stops paying for a server, and that has happened to a great deal of good equipment. This turbine has no such dependency. It will behave in 2045 exactly as it behaves on the day it is bolted down, whatever has become of us by then.

A connected turbine shows a pupil a number on a screen. This one makes them go and measure it.

That is the third advantage, and it is the real one. An app hands over a figure and hides everything behind it. A charge controller with accessible terminals is an instrument: a class can put a multimeter across it, log voltage and current against wind speed, plot a curve, and argue about why the curve is not the shape the brochure implies. You cannot do that with a sealed appliance. You can do it with this all day.

One of the forty, opened for this photograph. Turbine, hardware and the boxed controller, exactly as they were packed.

Free to schools. Half price to resellers.

If you are a school, a college, a university department or a training centre, the turbine is free. Not discounted — free. If you are a reseller, the same stock is available at half price. Either way, the route is the same: create or sign in to an account and contact us from there, so the enquiry arrives attached to a real address we can quote shipping against.

Please be direct with us about what you are and what you intend to do with it. Forty is not many. We would rather they went to forty places that will actually put them up than to four hundred that are curious.

What we are not giving you.

This is the part we would ask you to read twice, because a free turbine that nobody planned for is not a gift, it is a liability with a two-year warranty on it.

We are giving you the machine. We are not giving you the installation. On a school site that means, at minimum: somebody competent confirming the structure or the foundation can take the loads; whatever planning or landlord consent applies where you are; the electrical work done and signed off by a qualified person; a risk assessment; your insurer told; and the mast sited or fenced so that it is not in the middle of where children play. None of that is unusual for a school, but all of it has to happen before the crate is opened, not after.

And one more, which is the one that actually decides whether this works: somebody has to own it. A named person whose job includes checking the fixings each autumn and knowing how to brake the machine before a storm. If that person does not exist, the turbine becomes a seized ornament in three years, and that teaches a lesson none of us wanted to teach.

Before you ask us for one.

A windy country is not a windy playing field. National wind maps say very little about a site enclosed by a sports hall, a line of trees and a three-storey teaching block. If you have a term to spare, measure first: an anemometer on the proposed mast position, logged through a windy season, will tell you more than any map and makes a better piece of coursework than the turbine will.

If the answer comes back that your site is sheltered, we would rather you knew that before we shipped a crate across a continent. Tell us and we will say so. There is no version of this where we want a turbine standing still in a playground with our name on it. The full specification for the Atlas range is on the product page, and the figures there are the ones to design against.

The rest of them, still banded. Loaded the same way as any other order — the only difference is the invoice.

Forty machines that cannot be updated, cannot be tracked, and cannot be switched off from somewhere else.
We think that makes them worth more to a classroom than they ever were to us on a shelf.

Forty complete sets are available at the time of writing: turbine, blade set, and charge controller with dump load, boxed in 2023, unused and not refurbished, each carrying a two-year warranty. They are free to schools, colleges, universities and training centres, and half price to resellers. Installation, structural assessment, planning or landlord consent, electrical work, certification and insurance are not included and remain the responsibility of the recipient. When the forty are gone this offer ends, and we will say so on this page rather than quietly leaving it up. No customer is named here. The turbine specification is our own published data and is checkable on the product page.

Companies announce factories. The announcement is the easy part — a ribbon, a press release, a photograph of people in high-visibility vests. What actually changes what a turbine costs is duller than that, and it looks like this: one machine at a time, bought and installed in a workshop we own and run ourselves, taking one more step of the build in-house.

A lathe is not an announcement.

A CNC lathe turns metal. A part spins, a cutting tool moves against it under computer control, and what comes off is round, concentric and the same every time. The word that matters there is every. A skilled machinist can make one excellent part by hand. A program makes the two hundredth part identical to the first, on a Friday afternoon, by an operator who was trained last month.

In a wind turbine, that means the round things everything else bolts to: shafts, hubs, bearing seats, housings, the mating faces that decide whether a rotor runs true or shakes itself loose over a winter. These are not glamorous parts. They are the ones that determine whether the glamorous parts survive.

Stator cores, stacked and waiting. Parts we cut ourselves are parts we do not queue for.

Every step you bring in-house is a price you stop paying.

When somebody else makes a part for you, you pay for four things: the part, their margin, the freight, and the wait. You also accept their minimum order quantity, which means you buy five hundred of something because that is the smallest number they will run — and you carry the other four hundred and twenty on a shelf until you need them.

Bringing the step in-house removes the margin, removes the freight, removes the minimum order, and shortens the wait from weeks to the same afternoon. That is not a clever strategy. It is arithmetic, and it is the entire reason our prices are what they are.

A machine only pays for itself if it runs. One that works a day a week is more expensive than buying the part.

That is the honest catch, and we would rather say it than have you assume we have not thought about it. A lathe is capital sitting on a floor. It needs an operator who knows what he is doing, tooling that wears out, maintenance nobody enjoys paying for, and — above all — enough work to keep it turning. Buying machinery is only cheaper than outsourcing at volume. We buy it when the volume is already there, not to look industrial in photographs.

The volume that justifies the machine. Generators in housings, before assembly.

Owning the machine means owning the fault.

This is the part that gets left out of most manufacturing pages, and it is the part we care about most. When a supplier makes your part, a defect is a negotiation: emails, photographs, a credit note, a replacement in six weeks, and a customer who is still waiting. When you make it, there is nobody to write to. It is yours. You stop the line, you find out why, and you fix the process rather than the paperwork.

That is a harder way to run a company and a better one to buy from. It also only works if you measure. A machine that holds a tolerance is worth nothing unless somebody checks that it still is — which is its own equipment, its own routine, and its own cost.

Cutting the part is half the job. Proving it is still right is the other half.

What “local manufacturing” honestly does, and does not, mean.

Local manufacturing has become a marketing word, so here is what it does in practice. It shortens the distance between where a part is made and where it is needed. That means less freight, fewer border crossings, shorter lead times, and stock sitting nearer the person who ordered it. When a household wind turbine reaches you in days rather than months, that is why.

Here is what it does not mean. It does not make the product better on its own. A well-run factory a long way away will beat a badly-run one down the road, every time, and we are not going to pretend geography is a substitute for engineering. Proximity buys you speed and cost. Quality you have to build separately, and then keep proving.

And what we still do not make ourselves.

Nobody makes everything, and any manufacturer who tells you otherwise is being generous with the word “make”. We do not smelt aluminium. We do not draw copper wire, grow silicon, or wind our own bearings. We buy magnets, bearings, fasteners, cable and electronic components from people who are better at making them than we would ever be, and we would rather buy a good bearing than build a bad one.

What we do ourselves is the part where the design lives: cutting, forming, machining, winding, coating, assembly and test. That is where a turbine stops being a pile of components and starts being a machine that has to survive a gale on somebody's roof. It is also, not coincidentally, where the cost and the blame both sit.

The end of the line, which is the only place the argument can be checked.

The machine before this one, a laser cutter, two years in and entirely unremarkable. That is what success looks like here.

In a year that lathe will be an unremarkable thing in the corner of a workshop, covered in swarf, with somebody's coffee on the control cabinet.
The only trace of it will be a price that did not go up.

The factory described here is owned and operated by TESUP; the machine is our own purchase, and every photograph on this page is of our own equipment, our own production and our own shipments. TESUP has been in operation since 2018. No customer is named here, no supplier is named, and no order detail is published. We have not put a figure on what the machine cost or what it saves, because we would not be able to show you the workings — and a number you cannot check is worth nothing. The turbine specification referred to is our published data and is checkable on the product page.

What does tell you something is a long series of measurements taken the same way for a very long time. The Met Office keeps one, going back to 1884, and published its latest reading of it in July 2026. It says something more interesting than “Britain is getting hotter”, and the interesting part is not the part that makes headlines.

The warning that prompted this piece — and the five clear days behind it. Warning details: Met Office, redrawn rather than reproduced.

One warning is weather. A century of records is climate.

2025 was the warmest year in the UK series that begins in 1884. Four of the last five years sit in the top five. The most recent decade, 2016–2025, ran 0.51 °C above the 1991–2020 average and 1.33 °C above 1961–1990, and the UK has been warming at roughly 0.25 °C per decade since the 1980s.

A single record year proves very little on its own; records happen. What is harder to argue with is that they stopped being rare. When the warmest year in a 140-year series keeps arriving, and the runners-up are all recent, the series itself has moved.

Content supplied by the Met Office — State of the UK Climate 2025, published July 2026.

The winters got wetter. Then the summer ran dry.

Across 2016–2025 the winter half-year, October to March, was 3% wetter than the 1991–2020 average and 13% wetter than 1961–1990. Very wet winter months — ones with twice the 1961–1990 average rainfall — have roughly doubled in frequency.

And then, in the same decade, the total flow of England's rivers between March and August 2025 was the second lowest in a record starting in 1961. Lower than every major summer drought since, apart from 1976.

Both of those are true at once. The signal is not simply warmer, or wetter. It is less predictable.

That is the sentence worth taking away, and it is a duller one than the headlines. A country can be getting wetter in winter and running its rivers dry in summer in the same ten years, because what is changing is the distribution, not just the average. Planning around an average is exactly what stops working.

Blossom in London in early March. Not proof of anything on its own — but this is what a shifting distribution looks like from a pavement.

And this year, the Pacific has a say.

There is one more thing sitting over this particular year. The Met Office has declared an El Niño for 2026, and current guidance is that it will be a very strong one, with sea-surface temperatures in the central and eastern tropical Pacific running more than 2 °C above normal.

It is worth being precise about what that does and does not mean for Britain, because it is routinely overstated. El Niño is a Pacific phenomenon, and its effect on UK weather is indirect. It does not deliver a thunderstorm to Nottingham on a Wednesday in August. What it does is shift the odds for the months that follow: a higher chance of a milder, wetter and windier autumn and early winter.

Set that against the summer just gone, which was hot and dry — ten days above 35 °C in 2026, three of them reaching 37 °C. A parched summer followed by a forecast of a wetter, windier autumn is the same pattern as the rest of this page, arriving inside a single year rather than across a decade.

And it points at the same conclusion. If the coming autumn and winter are likelier than usual to be wet and windy, that is precisely the half of the year when a solar roof gives you least and a turbine gives you most.

Frost is disappearing, and that changes what a house does.

Two of the clearest numbers in the report are about cold, not heat. Compared with 1961–1990, the most recent decade has had less than half as many icing days — days that never rise above freezing — and hard frosts, below −8 °C, have fallen by a factor of more than four. At the other end, Greater London has had more than four times as many days above 30 °C and nights above 18 °C.

For a household that is not an abstraction. It is less heating in some winters, more cooling in some summers, and a demand curve that no longer looks like the one your house was designed around. We are not going to tell you whether that is good or bad for your bills, because it depends entirely on your house — but it does mean the shape of what you use is moving, not just the total.

Content supplied by the Met Office — State of the UK Climate 2025.

Sea level is the measurement that does not argue back.

Temperature records invite debate about instruments and station moves. Tide gauges are harder to talk your way around. UK sea level has risen by about 20.1 cm since 1901, with a likely range of 16.6 to 23.6 cm, and roughly two thirds of that has happened in the last three decades — 13.9 cm of it between 1993 and 2025 alone. The rate is increasing.

That matters to anyone on a coast, because sea level is the baseline a storm surge is added to. The same storm arriving on a higher sea reaches further inland. If you live near the water and you are choosing where to put anything — a turbine, a heat pump, a consumer unit — that is the number to think about, not the forecast for Tuesday.

Every centimetre of sea level rise arrives here first, and only later on the forecast.

So what does a turbine on your roof actually do about this?

Nothing measurable. We will not pretend otherwise. One household wind turbine does not move a national emissions figure, and if reversing climate change were the only reason you were considering one, you would be better off spending the money on insulation and voting.

What it does is narrower and more useful. If the thing that is changing is reliability rather than average, then the sensible household response is not a bigger version of one source — it is a second one that fails at different times. In Britain the windy half of the year runs roughly October to March, which is exactly the half that is getting wetter and stormier, and exactly when the days are shortest and a solar roof produces least. The two do not fail together, and that is the whole argument.

The caveats are the same ones we give everybody. A windy country is not a windy garden: the national maps say very little about the lee of your neighbour's house. Measure your own site through a windy season before you buy anything, from us or from anyone. Know how to brake the machine before a front arrives. And check the mounting every autumn, because the winters in that dataset are the ones getting wetter and windier.

A national wind map says almost nothing about the gap between two buildings, or the lee of a hedge in Devon.

The honest summary of British weather is not that it is getting worse. It is that it is getting harder to plan around.
Everything above is an argument for not depending on one thing.

Every climate figure on this page — the 1884 series, 2025 as its warmest year, 0.51 °C and 1.33 °C for 2016–2025, 0.25 °C per decade, winters 3% and 13% wetter, icing days and hard frosts, the London heat counts, English river flows for March–August 2025, and sea level at 13.9 cm since 1993 and about 20.1 cm since 1901 — is published by the Met Office in State of the UK Climate 2025 (Kendon et al., International Journal of Climatology, July 2026). The El Niño declaration for 2026, its expected strength, and the description of its UK effect as indirect and weighted towards autumn and early winter are also the Met Office's, as are the 2026 summer temperature counts. We have not attributed any single storm, warning or dry summer to climate change, because a single event cannot be attributed that way from a report like this one. The turbine specification referred to is our own published data and is checkable on the product page. No customer is named here.

Most places we ship to, the argument writes itself: cleaner power, lower bills, less coal. New Zealand is not most places. It solved that problem decades ago, with water. Which makes it a more interesting place to send a wind turbine, not a less interesting one.

Oamaru sits in the district New Zealand dammed.

Oamaru is the largest town in North Otago and the main town of the Waitaki District — about 14,000 people on the Pacific coast, roughly 120 km north of Dunedin. Inland from it runs the Waitaki, a river carrying eight power stations. The dam at Benmore was finished in 1965 and is still the largest earth-fill dam in the country; Aviemore followed in 1968, on a river already dammed at Waitaki in the 1930s.

This is not a place waiting to be shown what renewable energy looks like. It is a place that built it at national scale, in concrete and rock, two generations ago. Anything we send there arrives into that context.

A country with more weather than people, and a grid built to use it.

We are not going to tell you this saves the planet.

In 2024, 85.5% of New Zealand's electricity came from renewable sources. That is one of the cleanest grids in the world. Adding one household turbine in Oamaru does not meaningfully change a carbon figure, and we are not going to pretend otherwise. In a country running on brown coal, a turbine on your roof is a climate argument. Here it is not.

If carbon were the only reason to buy one, we would tell you to spend the money somewhere else.

There is a different reason, and in New Zealand it is a much better one.

The argument is the dry year.

A grid that runs on water is only as full as the lakes. In 2024 the country found that out again: after a dry May, June and July, hydro generation fell to 23,490 GWh — the lowest since 2013 — and the renewable share of electricity slipped from 88.1% the year before to 85.5%.

Water is a battery the weather refills. Most years it refills it generously. Some years it does not, and the whole country feels it at once. Living in the Waitaki does not exempt you from that — those lakes are national infrastructure, not your personal supply.

All figures from MBIE, Energy in New Zealand 2025, reporting the 2024 calendar year.

New Zealand's own answer has been wind.

The country did not respond to the dry year by building more dams. It built wind. Wind generation hit a record 3,919 GWh in 2024, up 22% on the year before, and installed wind capacity went from 691 MW in 2020 to 1,269 MW in 2024 — close to double in five years.

A household wind turbine is the same answer at the scale of one roof. Not a replacement for the grid, and not a moral position — just a second source that does not depend on how much rain fell in the Southern Alps this winter. And the timing lines up: the windy half of a New Zealand year runs roughly April to September, which is also when the days are shortest and the lakes are lowest.

Crated and loaded. From here it is roughly eighteen thousand kilometres and a change of hemisphere.

What we would say before you order one.

Oamaru is a coastal town, and the coast is hard on hardware. Salt gets into fixings and brackets long before it gets into anything you notice. Check the mounting at the start of every windy season, look at the mast and bracket for movement or corrosion, and treat the fixings as a maintenance item rather than a one-off.

Otago southerlies are not decorative either. Know how to brake the machine before a front arrives, and if you are somewhere that issues wind warnings, have the cover to hand and use it. Our six-blade Atlas set is rated 5–35 m/s and built for exposed coastal sites, but a turbine is never sounder than what holds it down.

And the part that matters most: a windy country is not a windy section. New Zealand has excellent wind on the national maps, and your own house may still sit in the lee of a shelter belt, a hill or a neighbour. Measure your own site through a windy season before you buy anything — from us or from anyone.

The only number that decides anything is the one measured on your own section.

New Zealand already generates its power from the weather. It just happens to have picked the weather that falls as rain.
We are sending the other kind.

Electricity figures — the 85.5% renewable share, the 88.1% for 2023, hydro at 23,490 GWh, wind at 3,919 GWh and 1,269 MW — are published by New Zealand's Ministry of Business, Innovation and Employment in Energy in New Zealand 2025, covering the 2024 calendar year. Population and geography for Oamaru and the Waitaki, and the dates of the Waitaki, Benmore and Aviemore dams, are from public record. No customer is named here and no order detail is published. Every photograph on this page is our own, of our own products and shipments. The turbine rating is our published specification and is checkable on the product page.

Almost every home energy decision gets made in summer. Long evenings, visible sun, panels at their best, and a quote that looks wonderful in July. Then October arrives, the days shorten, the bill climbs, and the roof goes quiet. It does not have to.

Wind and sun keep opposite calendars.

This is the part that surprises people, and it is not a marketing line — it is meteorology. In the temperate latitudes the wind is strongest across October to March, when the temperature contrast between pole and equator is at its sharpest and the storm track runs hardest. Sunshine does the exact opposite, peaking in June and July and bottoming out in the dark months.

The two resources are not competitors. They are out of phase. The weeks your panels give you least are the weeks the air outside is moving most.

Schematic. The shape of the year, not measured values — your own site decides the numbers.

Your bill keeps the same calendar as the wind.

Household electricity use is seasonal too, and it moves with the wind rather than the sun. Lights on at four in the afternoon. Heating and heat pumps working. More hours spent indoors, more cooking, more everything. In most of the countries we ship to, the heaviest domestic consumption of the year falls within October to March.

So the worst month for a panel is often the biggest month for a bill. A roof carrying only solar has been built for the half of the year when the bill is smallest. That is a strange way round, and almost nobody notices it, because the decision was made in July.

Build only the sunny half and you have built for the cheap half of your year.

What that means for one roof.

A battery does not care which machine filled it. Put a wind turbine and semi-flexible panels on the same roof, into the same battery, and the two curves fill in for each other: the panels carry the long days, the turbine carries the dark, windy ones. Neither has to be sized for the whole year on its own.

That is the entire argument for building both halves, and it is the reason we make both rather than making one and pointing at somebody else's.

Two machines, one battery, opposite seasons.

The windy season is also the storm season.

Here is the half of this that a seasonal sales piece would leave out. The months that generate best are the months that load your mounting hardest. More energy in the air means more force on the bolts, the mast and the roof structure underneath.

So October is the month to go up and check the installation, not just to enjoy the output. Torque the fixings. Look at the mast and the bracket for movement or corrosion, especially near salt air. Know how to brake the machine before a named storm arrives, and if you are somewhere that issues wind warnings, have the cover to hand and use it. Our blade sets are cut for real conditions — the six-blade Atlas set is rated 5–35 m/s and built for exposed coastal sites — but a turbine is only ever as sound as what holds it down.

The mast, the bracket and the pad underneath — the parts worth an hour before the windy months arrive.

A windy country is not a windy roof.

Everything above is the shape of the year in general. It is not a measurement of your house. A national wind map tells you about open ground at ten metres; your roof sits in the wind shadow of whatever is upwind of it — a taller neighbour, a line of trees in leaf, the ridge of your own building.

If you are near a coast, on a ridge, or on open farmland, the seasonal pattern above will probably show up on your roof much as it does on the chart. If you are inland in a dense street with mature trees, it may not show up at all. The only honest way to find out is to measure your own site through a windy season rather than trusting the map or trusting us.

Measure the roof you actually have, through the season that actually matters.

The sunny half is already covered on a lot of roofs. The windy half is the one still sitting unused — and it arrives in October.
Measure first. Then build the other half.

The seasonal chart on this page is schematic: it shows the phase relationship between wind, sunshine and household demand through a year, not measured values for any particular place, and we have deliberately put no numbers on it. Actual seasonality varies widely by latitude, terrain and local climate, and near the equator it is far weaker than shown. Every photograph here is our own, of our own products. The turbine figures are our published specifications and are checkable on the product page. For your own site, measure it — that is the only figure that decides anything.

TESUP is the global leader in household wind energy. Almost a decade of building household wind turbines and semi-flexible solar panels — both halves of a working system, engineered by one company, for one roof, one battery, one family. Thirty-four countries. Twenty-three languages. One standard.

Almost a decade in business. Half a century of engineering.

TESUP began active business in London in 2018. Almost a decade of designing, building and standing behind household wind.

The company is young. The engineering is not. Our expertise runs back to 1974 — through hydroelectric, wind and solar power plants totalling more than 1,000 MW of installed capacity. Utility scale. National scale.

That is what almost nobody else in this market has. Most competitors began with a small turbine and worked outward. We spent fifty years building the large ones, then brought that discipline down to something that fits on a roof. The scale changed. The standards did not.

Two halves. One system.

Almost every company in small-scale renewables makes one thing. Wind companies make turbines, then point at somebody else's panel. Panel companies make panels and have never machined a blade. Two industries, one customer, nothing shared.

One roof. A turbine and panels, working the same battery.

We make both. Household wind turbines and semi-flexible solar panels. One company. One system. One battery that does not care which machine filled it, only that something did.

The sun sets every day, everywhere, without exception. The wind keeps different hours entirely. Build only one half and you have built half a day. We build the day.

Anyone can sell you a machine. Only a company that builds both halves can tell you which one you actually need.

The other half. Semi-flexible — it bends to the surface, carries no frame, and weighs a fraction of glass.

Thirty-four countries. Twenty-three languages.

38 retail storefronts. 34 countries. A rest-of-world store for everywhere else. 23 languages. Each with its own currency, its own tax treatment, its own shipping, its own terms. Every one mirrored again for wholesale — 76 storefronts held in step.

Drawn from our own store configuration. Check it against the store switcher at the top of this page.

Global is not a word you earn by shipping abroad. It is what happens when a customer in Osaka, in Rio, in Riyadh and in Trégunc each open a page written for them, priced for them, in their own language — and none of them can tell which one we built first.

Cut in-house. Down to the millimetre.

Blade sets are laser-cut in our own building, from 1.2 mm high-strength aluminium, on our own fibre laser. Three profiles from the same stock on the same bed: 2–20 m/s for 2 kW peak. 4–25 m/s for 3.8 kW. A six-blade set at 5–35 m/s for 4.5 kW, built for exposed coastal sites.

Owning the machine means owning the shape. A subcontracted blade is a purchase order and a six-week conversation. A blade on our own bed is a file, a test cut, an afternoon. That is the difference between a company that assembles turbines and a company that designs them. Manufacturing runs mainly in Slovakia and London, and in Turkey as well.

And the part most companies leave out

Leading a market is not the same as selling to everyone in it. So we keep saying the things a smaller company cannot afford to say.

If your roof sits deep in a dense old city, walled in on four sides, do not buy a turbine from us. Buy the panel. If you are inland with no real wind, measure first, then decide. If you have a strong flat roof that will carry the weight, buy rigid glass instead of semi-flexible — for that job it is the better panel.

We can say all of it because we build both halves. When the honest answer is “not the turbine,” we still have the one that works. A customer told the truth in year one is the reason there is a year eight.

Fifty years of engineering. Almost a decade of building. Thirty-four countries. Both halves of the system.
That is what global leader is supposed to mean.

The founding year, the 1974 engineering lineage and the 1,000 MW installed-capacity figure are our own published company record. Every photograph on this page is our own, of our own products. The map is drawn by us from our own store configuration, and the storefront, country and language counts can be checked against the store switcher on this site. The turbine figures are our own published specifications and are checkable on the product page.

Where Trégunc is, and why it matters

Chantier naval Marée Haute sits at Trégunc, in the Finistère, on the south coast of Brittany between Concarneau and Pont-Aven. The yard is on the Minaouët, a short tidal river running down to a coast that faces open water to the south and west. The department is called Finistère for a reason: finis terrae, the end of the earth. It is the westernmost part of mainland France, and the weather arriving there has crossed the Atlantic without meeting anything first.

Trégunc on the south Finistère coast. Drawn by us from public coordinate data.

We have spent a lot of these posts telling people that their site is probably worse than they think. We told buyers in Palermo that a roof in the middle of the old city is in the wind shadow of everything around it. We told buyers in Saudi Arabia that most of the interior is not windy and that they should measure before they spend anything. This is the post where we get to say the opposite, and it feels strange enough that we want to be careful about it.

The work and the wind arrive in the same month

Here is the part that is specific to a boatyard rather than to Brittany in general. A yard's year is not flat. Boats come out of the water in autumn and go back in around spring, and the months in between are when the heavy work happens: hivernage, carénage, antifouling, aérogommage, sanding, painting, engine work, refits. That is compressors running, extraction running, lighting running, in a shed, in the dark half of the year.

And the dark half of the year on the Atlantic coast of Brittany is the windy half. The depressions that come through in winter are the strongest weather the region gets, and they arrive in exactly the months when a solar panel is producing least and a yard is drawing most. That is not a coincidence we engineered. It is just what a maritime climate does, and it happens to line up with how a boatyard works.

A boatyard year, drawn as three shapes. Deliberately schematic and not to scale — it is the phase relationship that matters, not the values.

Everywhere else we argue that a panel and a turbine fail on different days. At a boatyard they also succeed in the right months.

A hull and a blade are the same problem

Marée Haute has been at Trégunc for more than twenty years, and one of its main activities is building Django sailing yachts from end to end — the 7.70+, the 9.80 and the 12.70 — alongside the service side of the business: mechanics and repowering, keel and rudder work, composite and timber repair, rigging, electronics, winter storage. In other words they laminate their own hulls.

That is worth pausing on, because a hull and a turbine blade are the same kind of object. Both are a shape whose only job is to move through a fluid without wasting energy in turbulence, and in both cases a small error in the profile is invisible in a photograph and permanent in the finished part. We cut our own blades on our own laser for exactly the reason a yard laminates its own hulls: because the shape is the product, and you do not subcontract the product.

We would rather sell to people who can catch us out

A yard does not need us to explain corrosion to it. Antifouling, aérogommage, paint on the œuvres vives, sail-drive seals, sacrificial anodes — managing what salt water and salt air do to hardware is not a footnote in that business, it is the business. Which means that if the finish on one of our machines is not up to it, the people at Trégunc will know before we do, and they will say so.

That is uncomfortable and it is the point. It is easy to sell hardware to somebody who will not notice for three years. It is a good deal harder, and much more useful, to sell it to a workshop that will have an opinion by the end of the first winter.

What we would hold against us, in your position

Salt first, because it is the obvious one and we are not going to pretend otherwise. Coastal air corrodes hardware left outdoors, ours included. Everything we sell is designed to live outside, but designed to live outside and indifferent to twenty years of Atlantic salt are not the same claim, and only the first one is ours to make. On that coast a machine of ours will need looking at, the way everything else in a yard needs looking at.

Second, a yard is not automatically a good site just because it is near the sea. Yards have big sheds, and a big shed is a big wind shadow. Boats on the hard with their masts up are not nothing either. A turbine tucked down between two buildings on a coastal site can do worse than one on an ordinary exposed roof twenty kilometres inland. The coast helps; the layout decides.

Third, a yard's demand is not a household's. Compressors and extraction are a different load from a fridge and some lights, and one small turbine is a contribution to that, not an answer to it. We would rather write that down plainly than let anyone infer we are proposing to run a boatyard on a single machine. Measure the site, look at what the yard actually draws in February, and size it from there.

The map and the seasonal diagram on this page are ours; the map is drawn from public coordinate data and the diagram is schematic rather than measured. The photograph of the Finistère coast is a licensed stock image, not our own, unlike the factory photographs elsewhere on this blog, which are. The two films are Marée Haute's own work, embedded from their public YouTube channel and not re-hosted by us. What the yard does and how long it has been doing it are taken from its own published description of itself. The turbine figures are our own published specifications and are checkable on the product page.

Palermo from above. Count the metres of clear air between one roof and the next — there are barely any. That density is the whole problem in one photograph.

A city behind a 606-metre wall

Monte Pellegrino is an isolated carbonate promontory on the northern edge of Palermo. At 606 metres it is the highest point inside the city boundary, it closes the Gulf of Palermo to the north, and it is visible from essentially everywhere in the city. For our purposes it is something less romantic: a very large windbreak, sitting exactly where the sea breeze would otherwise arrive.

Behind it the old city occupies flat, dense ground, and behind that the land climbs again into the hills. A turbine on a roof in the middle of the old core is standing at the bottom of a bowl, in the wind shadow of every building around it, and roughly nothing on a specification sheet will rescue that site. Move the same machine to the coast, or up onto the slopes where a building stands clear of its neighbours, and it is a different proposition entirely.

The same city drawn as a section. Deliberately schematic and not to scale — the point is shelter and exposure, not metres.

One island, two extremes

It is worth saying that Palermo is only one end of Sicily. At the other, on the east coast some two hundred kilometres away, Etna stands at roughly 3,403 metres — the tallest peak in Italy south of the Alps, and a mountain whose exact height changes when it erupts. Against that, the highest ground inside Palermo's city limits is a 606-metre promontory. The photograph at the top of this page is the eastern end of that range: commercial turbines standing on an exposed Sicilian ridge with Etna smoking behind them.

An island that runs from a 3,400-metre volcano to a sheltered coastal plain does not have one wind climate, it has dozens. That is the real reason we will not quote you a number for “Sicily”. The honest unit here is not the island, or even the city. It is the site.

The wind that ruins the solar day

Then there is the sirocco: a Mediterranean wind that comes off the Sahara and can reach violent speeds by the time it makes landfall. It arrives over Sicily from the south, carrying North African dust with it, and for a solar panel it is straightforwardly bad news. The haze cuts the light, and the dust settles on the glass and stays there until somebody climbs up and washes it off. What is easy to miss is that this is also the strongest wind the island gets.

That is the useful thing about running a panel and a turbine on the same battery in a place like this. They do not fail on the same days. The hazy, dust-laden afternoon that flattens a solar curve is very often a windy one, and the still, glassy August morning that suits the panel perfectly gives the turbine nothing at all.

A panel and a turbine are not twice as reliable. They are unreliable on different days, which is the part that actually matters.

Why Palermo keeps coming up

Palermo is the capital and largest city of Sicily, a 2,700-year-old city on the gulf that carries its name, with something like 1.2 million people in and around it. It is also, quietly, one of the stronger parts of our Italian map — not because we have any romantic attachment to the place, but because that is where the orders come from. Once you know the geography that stops being surprising. A large coastal population, a great many buildings outside the old core that stand on their own, and a season of genuine wind add up to somewhere a turbine can earn its keep.

Cefalù, further along the same coast. Compare it with the first photograph: a town built out onto a headland with open water on both sides is the opposite of a sheltered courtyard, and it is the kind of Sicilian site where a turbine has something to work with.

What we would hold against us, in your position

The sirocco is seasonal and episodic. It is not a resource you can bank on every week, and anyone selling you a turbine on the strength of one dramatic wind is selling you the exception rather than the average. The dust that dirties a panel also finds its way into anything mechanical, and coastal salt air corrodes hardware left outdoors — ours included. Both are maintenance you will genuinely have to do, and we would rather write that down than let you discover it in year three.

And the honest summary of everything above: if your roof is in the middle of the old city, hemmed in on four sides, do not buy a turbine from us. Buy the panel. We would rather sell you half a system that works than a whole one where one half spends its life stationary. If you are on the coast, on the slopes, or on a building that stands clear of what surrounds it — measure the site, and then talk to us.

The section drawing on this page is ours and is schematic rather than surveyed. The three photographs — the turbines below Etna, Palermo from above, and Cefalù — are licensed stock images, not our own, unlike the factory photographs elsewhere on this blog, which are. The figures for Monte Pellegrino and for the city are the published ones, and the sirocco is what the meteorological literature describes: a Mediterranean wind out of the Sahara capable of very high speeds. The turbine figures are our own published specifications and are checkable on the product page.