LAST WEEK'S ANNOUNCEMENT that the Darlington nuclear power development will be the site of two new generating units is both interesting and unsettling.

Our interest lies in the government's apparent decision to try untested technology, by opting for two units that would be designed to produce almost as much electricity as now comes from the existing fourunit station.

Making his last announcement since his job was taken over by George Smitherman, Energy Minister Gerry Phillips said that, as had been widely expected, publicly owned Ontario Power Generation will own the new plant on Lake Ontario near Oshawa, construction of which will construct a new plant east of Toronto in a project which will create about 3,500 direct construction and engineering jobs between 2012 and 2018.

Three firms are bidding to design the power plants: the federally owned Atomic Energy of Canada Ltd. (AECL), which makes the existing Candu reactors; U.S.- based Westinghouse Electric, and Areva of France, the world's largest nuclear reactor company.

The Liberal government currently wants nuclear plants to remain the source of half of Ontario's electricity in coming decades as the province weans itself off heavily polluting coal plants and relies less on oil and natural gas.

Coal-fired plants, which contribute to deadly smog and greenhouse gas emissions, are to be phased out in 2014, well before the new nuclear units are in service, and it remains to be seen how the 2014 target can be met without exposing Ontario to A need to import huge amounts of power generated by even dirtier coal plants in the Ohio Valley.

As we see it, the course chosen by the McGuinty Liberals is unnecessarily risky, and may well commit the province to far too much reliance on energy imports.

The decision to have competing bids may well have stemmed from a desire to minimize costs, the government knowing full well that AECL historically won costplus contracts as the only entity whose reactors use the Candu technology.

As we see it, the OPG is certain to face a classic "apples and oranges" situation, with the two foreign firms able to submit relatively low bids for huge reactors that require enriched (weapons grade) uranium and AECL with no proven design of reactor capable of generating more than 1,000 megawatts.

A critically important consideration, beyond the area of safety and the potential need for uranium enrichment is the impact of the contract award on Canada's fragile manufacturing sector and in particular its nuclear industry.

Clearly, a decision to "save money" by going with the foreign firms will result in an enormous blow to our nuclear industry similar to the blow dealt our aeronautical industry by John Diefenbaker's killing of the Avro Arrow half a century ago.

As we see it, a vastly preferable approach would have been to invite proposals for an additional 4,000 megawatts while requiring designs that use heavy water and ordinary uranium and at least two-thirds of the manufacturing in Canada.

We also have misgivings on the siting decision. When one looks at the relative performances in recent years, it appears that Bruce Power, the private consortium that has managed the Bruce Nuclear Power Development (BNPD) for OPG, has done a far better job than OPG has itself at the Pickering nuclear complex.

Bruce Power, which just last week offered to build Saskatchewan's first nuclear power plant, seems to be succeeding where OPG failed so miserably in bringing mothballed nuclear units back into service.

Although only time will tell, Bruce Power currently seems confident it will be able to resurrect two 750-megawatt units at the Bruce A Generating Station, one of which has been problem-plagued since the day it went on line because of junk apparently left by construction workers in its nuclear boilers.

Unlike the situation in Pickering, where the main work was with the pressure-tube reactors, the Bruce A project involves replacing both pressure tubes and boilers.

In the circumstances, the best course for the McGuinty Liberals would have been to hedge its bets, by leaving OPG in charge of an expansion of Darlington while having Bruce Power undertake a similar-sized expansion at BNPD.

Ideally, the Darlington project would have been a simple "twinning" of the existing station - a concept Ontario Hydro employed successfully at both Pickering and Bruce - to add about 3,600 MW of generating capacity without major changes in scale or technology.

At Bruce, the ideal arrangement would be to have AECL and Bruce Power collaborate on building the first two "second-generation" Candu units, with a power output of perhaps 3,000 MW.

Yes, that's a lot of additional generating capacity, giving us a total output of nearly 20,000 megawatts when all 24 reactors (10 at Bruce, eight at Darlington and six at Pickering) were running. But the reality is that many of the units will routinely be shut down for servicing, and if governments and Canadian society are going to get serious about greenhouse gases, the demand for non-polluting electricity will be virtually insatiable.

When one considers the fact that Toyota has already sold more than a million hybrid cars, the planned advent of plug-in hybrids in 2010 will itself have an enormous impact on long-term power demands that will best be met by "base-load" nuclear generating units.