Wind a Driver for Ontario Electricity Export Losses

Posted by on

Simple linear regressions of hourly wind output vs. hourly net exports for a couple of time periods yield the following:

  • For all of 2010, the regression coefficient is 20.5% with an R Square of 4.2%.
  • For January 2011, the regression coefficient is 29.5% with an R Square of 8.7%.

The low R Square appears to be a function of high volatility in both wind output and net exports. (Note that Ontario is sometimes a net importer.) However, the regression coefficients suggest that a substantial fraction of Ontario’s loss making exports are driven by wind power.

Ontario’s average revenue from electricity exports in 2010 was 3.4 cents/kWh. The average cost of power was almost 7 cents/kWh and the cost of power contracted in 2010 would be at least 15 cents/kWh. (Please provide a better estimate if you have one.)

 

7 Comments

  1. Page 17 (page 18 of the pdf) of the most recent OEB Regulated Price Plan (RPP) Price Report (http://www.oeb.gov.on.ca/OEB/_Documents/EB-2004-0205/RPP_Price_Report_20101018.pdf) shows a graph of projected spot prices and GA costs. Given that most of the market electricity is contracted and the total market price is about 7 cents/kWh and rising, I’d say the weighted contract price IS 7 cents. The average blended FIT price is in the order of 15 cents. The contracted average is of course pulled down by OPG and the Bruce B floor price, then Bruce A is around 7 cents, then the other, non-FIT stuff and FIT come in above 7 cents. Concerning the OEB graph, the only thing I’m uncertain about is whether it’s describing what’s going on with the total MARKET quantity or the total CONTRACTED quantity. If it’s talking about the MARKET, then not sure how they arrive at their projected market RPP supply price (“total contract cost”) of 7.269 cents (page 3 of doc, 4 of PDF). ALSO, the GA reallocation, not added to their projection, gooses the numbers upwards, by close to 0.2 cents.

  2. I read the post differently – the cost of “at least 15 cents/kWh” was contracted, and I assume that means initially contracted (not renewed), during 2010. That would be tough to get at – I guess one could tally each source coming online and uncover it’s contracts as best they could (but I think even Halton Hills alone – the simplest of the bunch – would be tough to get at). A cynic might add the costs of canceling the Oakville plant to that too.
    The HOEP plus the GA thing is irrelevant – although there’s some good information in the OEB document – specifically, when 92% of your supply is contracted in advance, and 10% of it is exported, the only things the HOEP does is determine the size of the GA, and the cost of exports. Thus the RPP, TOU, and Wholesale rate plans all end up about the same, because they are driven by contracts, not markets.
    In the past, I have taken all revenues (Ontario Demand @ HOEP + GA, Exports @ HOEP, less Imports @ HOEP). Known expenses become OPG (their 2010 annual report isn’t out yet), and Bruce. That leave you with $X to calculate for Y MWh. This document suggests adding $400 million of the NUG GA allotment to OPG. But it won’t be 15 cents (around 11 for 2009 if I recall correctly)

    Regardless, what the data is showing more clearly than relating wind directly to exports is relating wind directly to the HOEP – the windier it is, the cheaper the HOEP rate. The cheaper the HOEP rate, the more expensive it is to export (without the safety net of the GA). So it’s all well and good to say we export 800MW every hour regardless of wind – and when it is windy we only step up exports 30-40% of the wind production, but when it’s windy the first 800MW of exports are cheapened by the wind.

  3. Scott,

    Not a biggie but I’d argue that HOEP + GA IS relevant. First, let’s set aside the cost of CDM (~ $ 400 million per year or about 0.3 cents/kWh). Now, let’s assume the contracts underlying the GA have market is pretty much completely hedged — the OEB’s says it’s 92 % hedged and an analytical argument can be made that it’s higher than that. So, under those circumstances, if HOEP went to 7 cents then the GA would go to zero. If the GA is zero at HOEP = 7 cents then that is suggesting the weighted price of the underlying contracts is 7 cents. This is the way a contract-for-differences (CFD) works in commodity markets such as Ontario electricity. The analysis is more complicated than that but that’s my view of things.

    Bruce

  4. I’d also suggest it’s worth looking at the seasonality of the relationship between wind and exports. Looking at data for Jan10 – Jan11, I calculated the correlations for rolling hourly periods 31 days long. The average correlation of the windows (ending Jan 31, 2010 to Jan 31, 2011) was 7.2 % but it also went as low as – 39 % and high as 41 %. The standard deviation was 17 % and the min and max were in the ballpark of the mean -/+ 2 standard deviations, suggesting to me that the data approximated a normally or perhaps lognormal (slight skew to the lower side). I didn’t graph the data but it might be interesting to compare it to Tom’s work on seasonal characteristics of wind.

  5. (corrected typos, Tom please delete my previous comment)

    I’d also suggest it’s worth looking at the seasonality of the relationship between wind and exports. Looking at data for Jan10 – Jan11, I calculated the correlations for rolling hourly periods 31 days long. The average correlation of the windows (ending Jan 31, 2010 to Jan 31, 2011) was 7.2 % but it also went as low as – 39 % and high as 41 %. The standard deviation was 17 % and the min and max were in the ballpark of the mean -/+ 2 standard deviations, suggesting to me that the data approximated a normal or perhaps lognormal (slight skew to the lower side) distribution. I didn’t graph the data but it might be interesting to compare it to Tom’s work on seasonal characteristics of wind.

  6. Bruce, I don’t disagree that there are ways the GA is relevant – but I think it’s biggest utility is in measuring the level to which the market is dysfunctional.
    CAMECO just announced the 2010 Bruce earnings down, due to 9% lower prices – with another reduction forecast for 2011. OPG’s unregulated hydro assets also get hammered by artificially lowering the HOEP through the frivolous contracting the GA mechanism has allowed.
    The arguments that wind is causing the oversupply, and with it both the elevated, and costly, exports, plus increasingly a lower price, should be being made by the IESO. If it’s not inherent to wind production, the IESO is doing a poor job integrating the OPA’s contracted output into our supply.

    The data, in my opinion, looks like the IESO became less capable, in 2010, of integrating wind to meet Ontario demand – with the oversupply resulting both in exports and depressing the HOEP. The linked spreadsheet shows the relationship:

    https://spreadsheets.google.com/ccc?key=0AkOMoLvcg5SIdF9aemp6RGw1MmdwT05hcVdYSzU5UXc&hl=en&authkey=CKbwzdUJ

  7. Scott, I agree on all fronts. I took a quick look at your spreadsheet and when I have more time would be interested in discussing it with you.

Comments are closed.