It is not a job for the faint hearted. A 500-tonne tower the height of a 10-storey building fitted with two 16m-diameter rotors is dropped into a Northern Irish loch for the first time and switched on in the belief that it will produce electricity from tides flowing through the loch. There can be no dry run – a thing like that is too expensive to rehearse and the kit is too big to mess about with. Yet three years down the line the turbine, known as SeaGen, is working, producing enough power to have made £200,000 (€230,000) per year so far – just as well since the whole scheme has stacked up at least £35m (€40m) in costs.

SeaGen – the brainchild of Peter Fraenkel, technical director and co-founder of Marine Current Turbines, the company running the project – is considered the first device to produce and continuously sell electricity from a tidal race. Fraenkel, who calls himself “a very optimistic person”, certainly has chutzpah. Undeterred by the expense and risk of installing one turbine, the company, of which he is technical director, proposes to install four or five in a row, probably in Scotland. They have to push ahead, of course, because otherwise they will never make any money. If planners give the nod, the company’s first commercial tidal array could be up and running by 2014. A planning decision is expected in the next few months.

If the whole venture looks odd, it is with good reason: it’s not unlike creating a jumbo jet before you’ve produced a biplane. For a start-up company’s first real commission, it all looks far too expensive and too massive a project, working in a medium Fraenkel himself describes as a “horizontal waterfall”.

Oceans and their sheer power and volume are the simple reason behind this – a force Fraenkel has had to contend with through more than 30 years working in water, from the Nile to the Scottish Highlands. However, he does admit to being surprised at the turn of events.

“Until two or three years ago I thought it looked good but I didn’t expect it to come to any kind of conclusion – I thought there might be a showstopper of some kind. But it just gathered momentum,” he says.

It’s true that he had experimented with smaller models of the turbine previously – as with any start-up, proof of concept was needed to get initial funding, and this was done using a 3.5m-diameter, 10-tonne rotor in a Scottish loch. Later, a larger turbine was tested off the coast of Devon in England. Fraenkel didn’t take any massive leaps as he developed the turbine.

Coming from an aerospace background and using his engineering expertise, he reasoned that some of the characteristics of aircraft propellers, which had already been designed into wind turbines, could be used underwater, for SeaGen is effectively a type of submerged turbine not so different from those seen in onshore wind farms.

One of the patented features of his invention for the tidal sector lies in its ability to change the position of its rotors, which both makes it more efficient and means it can come to a halt gradually, creating less strain on the machine.

To the layman, that is not an obvious switch. Turbines above water or onshore are hard enough to work with. The waters at Strangford Lough, County Down, can flow at nearly 5m/s, compared with 1.5m/s in Egypt, where he first produced a turbine to drive an irrigation pump. Water itself is thicker and much more powerful than most winds. Its velocity is a key to the amount of electricity a turbine can produce, but by the same token that makes it a devil to work with. “Colleagues have worked out that if you put SeaGen on a mountaintop, you would be dealing with the equivalent of winds going at 250 miles an hour – twice the speed of a hurricane,” says Fraenkel with a wry smile.

So it is with good reason that the company has decided to produce this giant, made of a number of steel, carbon-fibre and other materials, planted 9m into the seabed. “The rotor blades have a force of about 30 tonnes (the weight equivalent of three double-decker buses) trying to bend them, so we had to use an extremely strong (but costly) material like carbon fibre to carry such a high load through a slim blade root,” explains Fraenkel. Divers have only half an hour or so during the dead time between tidal flows to carry out any minor maintenance checks; after that, they would be dragged away.

Repairs are a nightmare but a key problem all tidal developers have to build into their plans. The biggest problem of all is installing the device in the first place – especially for the first time in a new location. “Installation at the moment is two-thirds of our costs and that’s unacceptable,” states Fraenkel.

Installing SeaGen was fraught with problems none of the team could have foreseen. For one thing, they had to wait for weeks for a delayed barge to arrive, costing a fortune. For another, plans to pump liquid concrete to fix piles pinning the turbine to the sea floor were wrecked by the viscious flow of the water, which sucked it out again. Unique installation mechanisms have been devised to deal with some of these problems.

Having hopefully learned from these projects, the company plans to repeat the whole business again and again until costs fall. An incremental approach to innovation may be one of the secrets behind its success thus far, allowing lessons to be learned little by little over years, even decades. It is for that reason that Fraenkel ridicules most of his rivals’ weird and wonderful turbines. There are dozens of other ideas in the field, ranging from devices shaped like reversed sails to those that look like underwater kites.

“The less innovation the better, so far as we’re concerned,” he explains, “because to minimise the technology risk you need technology that so far as possible is tried and tested. We use fairly conventional power units (gearbox, generator and power electronics) and conventional rotor-design methodologies, whereas competitors with numerous innovative features are having serious difficulties in getting any results.”

SeaGen itself has been criticised for being too big and too costly. But many smaller devices have their own disadvantages, since they create less electricity and work in slower waters, yet may incur fixed and relatively high installation and connection costs. They may also be less able to carry some of the internal electronics and cabling necessary to connect arrays to the onshore grid in a cost-effective way.

From an ecologist’s point of view, SeaGen seems to be scoring top marks. Before installing the turbine, there was no knowing how local seals – which could be sliced in two by a rotor – would react. But these have sensibly avoided the turbine, as Fraenkel says he foresaw.

If the whole scheme sounds like a black hole for investors’ money, it is not surprising. It is an offshore infrastructure project, and there are not many of those across the world except in oil and gas. The leaders of these multinationals also took huge risks to build rigs in the deep waters of the North Sea in the 1970s, but with more of their own cash and direct government assistance.

Marine Current Turbines, on the other hand, employs 21 people and relies on dedicated supportive partners such as Germany’s Siemens, as many venture capital companies will not touch such a long-term programme. The government gives R&D assistance and a helpful mechanism for electricity payments.

Dr Andrew Tyler – the company’s new chief executive, recruited this summer as it moves through to full commercialisation – is looking for new investment. “We need companies that are original equipment manufacturers (OEMs) that are likely to invest as part of a long-term strategic intent,” he says.

For Fraenkel, the most challenging aspects of the whole project appears to be behind him. He argues that the next phase, building an array in a similar loch, will be easier. “We won’t make the same mistakes with the second, third and fourth turbines and we’ll do them in less time. Overheads will be split between turbines so there’ll be economies of scale,” he points out.

Is it worth all the trouble? Climate change policies aim to slash carbon emissions in the next 20-30 years so tidal stream energy could contribute to that. Still, there are other ways that work more quickly, such as insulating buildings.

It is a remarkably bold scheme, which could eventually create new jobs. According to Fraenkel, “The core business of delivering and installing tidal turbines becomes ‘revenue neutral’ – that is, capable of trading without recourse to further finance – by about 2015 if we can get moving rapidly.” Whether it develops into a niche market or becomes a commonplace form of energy is anybody’s guess.