Monday, November 2, 2015

October revealed flaws in Ontario's rate and electricity market designs

October 2015 provided a unique opportunity for analysis of Ontario's electricity supply system as the Darlington nuclear power station was unproductive due to a planned vacuum building outage. During the previous October Darlington generated 2.5 million megawatt-hours (MWh) of electricity, which was 23% of what the province consumed during the month. 2015's October therefore provides a glimpse of what Ontario's sector will be like in a few years, as the Clarington Transformer Station is completed and Pickering Nuclear Generating Station is closed.

Vacuum building outages are scheduled for the lowest demand periods of the year, and October is one of those month. October 2014 saw a glut of power and concluded with a $1+ billion estimate of the global adjustment (the difference between what the system pays suppliers and what is recovered through the sale of power at market prices). With the rise of solar (largely unreported in the province) Ontario regularly exceeded $900 million during the sunnier months of this year's second quarter, but 12 months ago $1 billion was a stunning, unprecedented number. Parker Gallant and I published an analysis of 2014's October noting:

  • the record high global adjustment total
  • the record global adjustment rate for the 90% of consumption that can be considered "Class B"
  • the record low Hourly Ontario Energy Price (HOEP) the market determined
  • the record low price for exports
  • the high level of curtailed generation (contracted supply the system could not take)
  • record supply from wind and solar generators.
Figure from IESO June 2015 18-Month Forecast
Ontario's system operator, the IESO, would include a graph of curtailments in June 2015 that showed Parker and I were close, if a little enthusiastic, in this claim:
During October, 2014, the IESO curtailed more than 500,000 MWh of production. While wind power accounted for 100,000 MWh directly, much curtailment at nuclear and non-utility generators (NUGs) and import cuts occurred due to bloated supply levels during periods of windy weather. If one combines the curtailed production with the exports for October, it is obvious that Ontario dumped more than 21% of the province’s procured (and paid-for) supply
October 2014 exports totaled 1.8 million MWh, dumped at an average price of about half of one cent per kilowatt-hour ($5/MWh). The 2.3 million MWh total, from exports sold at an enormous loss and supply simply curtailed, is only slightly off the 2.5 million MWh reduction in Darlington's production the following October. Ontario's demand was little changed from 2014's October to 2015's.[1]

It might surprise readers that much of Darlington's missing generation was replaced with generation from other sources, as exports dropped only 300 thousand MWh.
  • Industrial turbines had a record month of production with the generators the IESO reports show reporting locations up 440 thousand MWh over the previous October. For the first time ever this set of generators produced over 1 million MWh, and the hidden distribution (Dx) connected turbines likely delivered another 150 thousand MWh.
  • Imports from Quebec  surpassed 1 million MWh for the first time ever - up 800,000 megawatts from October 2014 and 65% higher than the previous monthly high for imports from that province
  • Ontario's natural gas-fired generators increased output by 800,000 megawatts.
  • Solar generation likely increased by 100,000 megawatts. [2]
Stitch all that together and you find a great deal of Darlington's generation was replaced. 

Without rewriting my work in Ontario is the sucker of first choice for off-price electricity, Ontario's contracting of supply is designed to depress Ontario's market price, making the IESO version of a market very attractive to exporters - including those making transactions that wheel Quebec electricity through Ontario to U.S. consumers.

The changes in the system's generation costs will look much different this year.

My estimated costs don't always match the costs the IESO passes on to consumer via the global adjustment - nor should the two necessarily match as I attempt to include all charges in the month the occur (including estimated Dx generation cost) and I omit costing curtailed supply costs. That being said, in my costing supply is not cheaper in October 2015 than one year earlier - but in the IESO's world, based on their 2nd estimate of the global adjustment for the month, it is. 

View the following graphic illustrating change in supply costs in tandem with the above change in supply - and you'll see that replacing nuclear with gas and imports may have reduced costs, but replacing with wind and solar cancels out any prospective benefit to the consumer.

This October's weighted average Hourly Ontario Energy Price (HOEP) rose to $25/MWh from last year's record low of $7.[3] As Parker Gallant and I displayed 12 months ago, the lower HOEP shifts costs from high consumption, primarily industrial consumers, to smaller "Class B" consumers.[4] My estimates:
  • Exporters' rates rise about 350%, but from a starting point of $5/MWh that just pushes it to $23/MWh
  • Class A consumers see the commodity rate (HOEP + Global Adjustment) up over 20%, to ~$62.50
  • Class B consumers see the commodity rate drop 9% from 2014's October, to ~$98/MWh
Overall there is a slight 2% decline using available IESO figures, while my supply based estimates show a slight increase. I've argued often Ontario has a low HOEP intentionally, as an industrial pricing policy. October 2015 shows a rising HOEP benefits smaller consumers.

There's another questionable pricing policy revealed in the month's data.
I noted from my preliminary monthly reporting page the average HOEP in the Ontario Energy Board's (OEB) Mid-Peak hours was $9.65/MWh higher than the average in On-Peak hours. That's not unexpected as the OEB reverses Mid-Peak and On-Peak hours Nov. 1 (until April 30). However, with the growth of solar it should not be a surprise mid-day pricing is not as different than other hours any longer, and the price premium does move to the starting hours of the day and the early dark hours. Similarly, I note in April Mid-Peak HOEP was higher than On-Peak. The data is suggesting the OEB move the start of winter pricing hours up to Oct. 1st and end it a month earlier, on April 1st.

From another perspective the most important figure for the month may be that greenhouse gas emissions likely more than doubled from 12 months earlier along with the output from natural gas-fired power plants. That number would be relevant only for Ontario because much of that generation was for export to states that would either have displaced coal or gas, but as Ontario strangely considers joining a California cap-and-trade scheme they might be cognizant the IESO's low-price policy encourages running gas-fired power plants for cheap exporting.


1. although the IESO will show it as down in some reporting, figures accounting for embedded generation will indicate actual consumption is nearly identical.

2. estimating solar production is not something one obsessed with accuracy should undertake, but we do know the one 100 MWe (AC connection) solar generator in IESO reporting produced 12,000 MWh during the month, and we know embedded solar is about 18 times that capacity now (so estimate 216,000 MWh). That's not my methodology, but my methodology does yield a similar 235 thousand MWh for October 2015, so it's a quick way to get to a similar number. My number for October 2014 was 135 thousand MWh.

3, what Ontarians pay on every MWh as a "Debt Retirement Charge" in an alleged attempt to pay off an alleged debt

4. Regulated Price Plan consumers essentially include most residential consumers - the distinction is the price plan rate is a prediction of what the Class B rate will be, whereas specifically "Class B" consumers will have the rate vary by month.

No comments:

Post a Comment