Source   : SaskPower 'Fact Sheets' for  hydro  and  wind

Source: SaskPower 'Fact Sheets' for hydro and wind

SaskPower's web page has short summaries on the generation options available for Saskatchewan. Their hydro page states it has 'high reliability' while the wind sheet says (you guessed it) it has 'low reliability'.

These statements are not however supported by the facts. From season to season hydro is about twice as variable as wind and there is no evidence this variability costs less than short-term wind variability. As it happens evidence suggests wind variability is significantly cheaper to manage.



Comparing seasonal variability in Saskatchewan's hydro and wind

First a word on capacity factors. Both hydro-electric power stations and wind turbines are different from a nuclear, coal or gas-fired power station because they have zero fuel costs -  both the water and the wind are free. For this reason whenever there is sufficient wind or water for them to operate - they will be used while more expensive coal and gas will be turned off. 

As a result we can assume that whenever wind and hydro CAN generate electricity - they WILL produce it. In other words they will generate the maximum amount of electricity possible given the availability of wind and water.

The measure for how much a generator has operated over a period is its 'capacity factor'. This is the actual output of the generator divided by the maximum possible for a specific period of time (usually one year). If a unit runs 100 percent of the time it has a capacity factor of 100 percent. If it never runs it has a capacity factor of 0 percent.

Examination of Saskatchewan wind and hydro capacity factors therefore provides a useful proxy for the availability and hence reliability, of these renewable resources.

Source: SaskPower Annual Reports (2009, 2012 & 2015)

Hydro-electric generation.  Over the period, with minimal change in installed hydro capacity (854 to 889 megawatts), hydro-electric generation varied between a low of 2,962 gigawatt hours GWh (2009) and a high of 4,641 GWh (2011).

This variation is due to fluctuating rainfall from year to year and is not cheap to manage. The difference between the two extremes (1,679 GWh) is equivalent to the output of a 370 megawatt gas-fired power station: the capital cost of such a facility would be around $750-million. Another way to consider it: the reserve requirement is 43 percent of the installed hydro capacity.

Wind variability. A simple visual inspection of the capacity factor chart above reveals the output of wind turbines is much more stable. Another way to present the data, which may make it easier to compare to annual hydro variability, is to assume the installed wind capacity is as much as is required to generate the same amount of electricity as the annual average produced by hydro over the period. The required amount of wind works out to be 1,240 megawatts. BTW: This is more than hydro (860 megawatts) due to wind's lower capacity factor.

The following chart shows, using the actual capacity factors for wind and hydro, how much that capacity would have generated since 2006 and,specifically, how much that annual output would have deviated from the average (the statistics wonks out there will note this is - sort of - a visual standard deviation);

Source   : SaskPower Annual Report  (2010 and 2015)

Source: SaskPower Annual Report  (2010 and 2015)

The chart illustrates very clearly that the variability of  hydro, from year to year, is significantly more than it is for wind. Over the 10-year period the maximum deviation of hydro below the average was 1,094 GWh - more than twice the 426 GWh for wind - and the maximum hydro deviation above the average was 602 GWh  or 55 percent more than the 389 GWh peak for wind.

Short- and long-term variability. Critics will note that, in the short term wind, is much more variable than hydro-electricity. This is absolutely true however the important thing is not the amount of wind's short term variability but how much it costs to deal with. An answer to that can be found in the recent General Electric study, profiled in our 18-July blog. It noted, among its other findings:  "the Canadian power system, with adequate transmission reinforcements and additional regulating reserves, will not have any significant operational issues if 35 percent of its electricity is provided by wind turbines."

The study itself went on to examine the cost of the additional short-term variability caused by 35 percent wind - in power systems speak - the need for additional 'regulating reserves'. Page 25 of the Summary Report "Regulation reserve requirements to mitigate wind variability appear to be a small fraction of the additional installed wind capacity...Overall the additional regulation across all of Canada was less than 1.7 percent of the installed wind capacity across all scenarios" [The maximum scenario studied assumed that wind provided 35 percent of total annual nationwide electricity needs.]

Experience from the United States

Source: US Energy Information Monthly - multiple editions since 2007

For anyone wondering - the experience South of the Border  with wind and hydro variability has been similar: this chart shows the rolling twelve month average of wind, solar and hydro-electric output in the US since 2007.

Wind and solar show rapid growth over the period nonetheless, and even without getting into any statistical analysis, it is clear there is significantly less inter-annual variability of wind and solar. Meanwhile hydro's annual output, with almost no change in installed capacity - around 100 gigawatts over the period - varies between a low of 235 terawatt hours (April 2008) and a high of 322 terawatt hours (December 2011).  



So there it is: a multi-year General Electric study, on generating up to 35 percent of all of Canada's electricity using wind, reveals the additional cost of regulating reserve for wind energy amounts to 1.7 percent of the installed wind capacity. Meanwhile the additional capacity required to make up for hydro variability in Saskatchewan, over just 10 years, amounts to around 40 percent of the province's installed hydro capacity.

The implication (and the finding from multiple electrical studies across North America and around the world) is that wind and hydro are ideal complements. Hydro can be used to smooth out daily variations in wind output while wind can be used to smooth out seasonal hydro variability.

So maybe it's time for SaskPower to re-think its statement about wind vs. hydro reliability.