THE MIND BOGGLES at the prospect of building a single wind power project in Lake Erie that's more than 30 times the 132- megawatt capacity of the Melancthon Wind Farm, currently Ontario's largest such project. Yet that's what the Dufferin project's owner, Canadian Hydro Developers Inc., now plans to accomplish.

Canada's largest independent developer of wind-energy projects is acquiring rights to the 4,400-megawatt "offshore wind prospect" from Utah-based Wasatch Wind Inc. The nearly 900 wind turbines to be erected in the shallowest of the Great Lakes would produce enough power at peak to meet the current needs of about 2 million homes.

Although the prospect might be welcome news at Queen's Park, where the governing Liberals have been pushing hard to have more wind farms built in the province, it raises some serious questions as to how best to meet Ontario's long-term electricity needs.

At present, the provincial government stands committed to phase out all coal-fired power production within the next five years and at present has no plan to invest in any new nuclear power plants. And it will be at least 2014 before the first stage of the Lake Erie project would be on stream.

Although it's clearly a "green" form of energy production, wind power has serious drawbacks in the area of predictability.

Historically, predictability was a problem for Ontario only when it came to predicting long-term needs. A failure to predict the surge in power demands after the Second World War led to a severe power shortage, and in the 1970s the failure of demand to meet expectations led to just as serious an over-supply of generating capacity and the need to cancel some projects and slow the construction of others.

With wind power, the real problem is the inability to predict a project's output beyond a few days because of the vagaries of meteorology. And even the largest wind project will produce little or no power on a hot, humid day when the demand for electricity peaks but there's nothing more than a slight breeze.

In the circumstances, there should be no doubt that Ontario's long-term power needs should be met by a sophisticated combination of base-load and peaking generators.

As we see it, any viable plan should include at least two new nuclear projects, conversion of the remaining coalfired plants to use natural gas, and the strengthening of inter-provincial grids to permit large-scale imports of power produced in Manitoba, Quebec and Labrador.

Instead of carrying out its plan to close the 4,000-megawatt Nanticoke Generating Station on Like Erie, the McGuinty government ought to set in motion the progressive conversion of its eight units to natural gas.

Although a few years ago such a conversion would not have made much sense economically, natural gas being so expensive and supplies being deemed so limited, that situation has changed dramatically with the discovery of huge untapped resources in shale deposits, not to mention the proven deposits in the Canadian Arctic.

And it just so happens that a lot of natural gas can be stored naturally in the Lake Erie basin.

Since the Hydro One transmission grid already provides for 4,000 megawatts of output from Nanticoke, it would seem fairly logical that the combination of the Lake Erie wind farm and conversion of Nanticoke to gas would leave the province with a new type of base-load capacity that would shift from wind to natural gas depending on the wind velocity.

The argument for two new nuclear plants would be based in part on the economic benefits to Canada of being able to prove anew the superiority of Candu technology in terms of safety as well as reliability.

However, any twinning of the 3,600- megawatt Darlington nuclear plant and addition of a new-generation Candu plant at the Bruce Generating Station would clearly require a risk-sharing agreement between the federal and provincial governments similar to those involved in the pioneering Douglas Point and Pickering A stations. (Such agreements would limit consumers' exposure to cost overruns and poor performance.)

As for the interprovincial transmission grids, there should be little doubt that over the long haul the best means of keeping Ontario's retail electricity prices competitive with those of other jurisdictions would lie in long-term contracts for power from new hydroelectric projects in Labrador and Northern Manitoba. But the challenge will be to get that power to Ontario consumers safely and economically through use of the latest transmission technology.

Of course, the one big unknown is the future level of demand for electricity, given the uncertainty relating to the provincial economy and the conflicting impacts of power-saving technologies and breakthroughs in battery technology that would confirm the future of electric vehicles.

Clearly, wind and solar will have a role to play, but both have their limits.