The press release for the OntarioEnergy Board's Market Surveillance Panel (MSP) Monitoring Report on Ontario'sElectricity Markets noted four recommendations, including: “an increase in the frequency with which interties are scheduled, and the associated frequency of demand and intermittent generation forecasts as well as pre-dispatch schedules” and; "accelerating efforts to make wind generators dispatchable ...” The MSP report's information supports the claims (argued previously here), that the depressed market prices, and dumping of excess generation, will not simply continue, but grow over the next decade if Ontario's supply procurement policies persist.
The Market Surveillance Panel (MSP) report delivers an analysis of “low-price” hours (defined as the Hourly Ontario Energy Price – or HOEP – being below $20/MWh). In comparing the Ontario Energy Board (OEB) seasons, where winter begins November 1st and summer begins May 1st, the report notes;
“The greater frequency of low-price hours in this year and in the past two years mirrors the general trend of lower Ontario demand and also reflects the increase to Ontario baseload supply, or generation that is offered like baseload supply (i.e. generators with fixed price contracts per MWh delivered). “
Baseload is difficult to define precisely in Ontario, as it has come to mean the electricity that the market needs to distribute because it is pre-purchased, regardless of demand. That generally includes nuclear units, a significant share of hydro generation capacity, a variety of, mainly natural gas, non-utility generators (NUGs), and generators contracted by the Ontario Power Authority (OPA) where the contract requires the grid to take all output (primarily wind and solar supply). Table 2.8 of the MSP Report (page 103) demonstrates the supply/demand relationship when the HOEP drops below $20/MWh. I've recreated much of that table, from existing data I've collected from available IESO sources; adding columns for wind data, and correcting a mistake in the April data on the original MSP Report.
|Year||Month||Number of Low-Price Hours||Ontario Demand||Net Exports||Ontario Demand Plus Net Exports||Wind Production||Wind Capacity Factor||Corrected MSP Excess Low-Priced Supply||MSP Ontario Demand Plus Net Exports||Corrected MSP Total Supply|
Presumably the “excess low-priced supply” was either curtailed by steam bypass (primarily at Bruce); by taking nuclear reactors, or NUG suppliers, off-line; or, just as likely, by allowing water to flow freely instead of diverting it through hydroelectric turbines (the tools available to the IESO are shown here). The MSP Report notes that frequently the ability to curtail wind production would eliminate some of the negative, and low, pricing. What caught my attention here was the high wind production during the low-priced hours. It appears that during low-price hours the recently contracted wind production was slightly higher than the low-priced supply being curtailed.
I queried IESO data by grouping on Hourly Ontario Energy Price (HOEP) ranges, and the results demonstrate some interesting characteristics of Ontario's dysfunctional market. As expected, low price periods feature high wind output, and high price periods don't.
Wind is particularly problematic due to its erratic intermittent character, but it's really the proportion of supply that the market is required to take, in scale to the overall market, that primarily determines the HOEP. Theoretically, a healthy market features a uniform price which is set by the last unit purchased. That price can only be meaninggully set where there is price competition, among suppliers. In Ontario market, the HOEP price is meaningless to the contracted generation - which is mostly referred to as "baseload" here. Adding to the confusion, some supply that is not run constantly, or on a must-take contractual basis, also has set prices; Ontario Power Generation (OPG) has regulated hydro assets meant to balance the grid and meet ramping demands throughout the day. I am grouping the data by baseload generation as a percentage of Ontario's demand, not inclusive of all hydroelectric output. That makes my measurements conservative, and the data shows prices dropping, and net exports increasing, at slightly above 85% baseload to demand. As baseload exceeds 100% of demand, we expect the price to be low, or negative (depending on export markets) - these situations are known as Surplus Baseload Generation (SBG) periods to the system operator that predicts upcoming SBG conditions, in order to plan supply curtailment actions.
The data for the OEB winter period of 2011 (November 2010 – April 2011) indicates the expected result when baseload is over 100% of Ontario demand. There is a surprising result at the other end, as net exports rise, probably as prices rise above the cost of fuels Ontario's suppliers found the capacity to fill export demand. The elevated net exports, when baseload was low and price higher, was not apparent in 2011's summer, indicating the capacity required to meet the peak demand of the summer is utilized to generate electricity for exports in the winter. Throughout the year, there is a consistent, inverse, relationship between low price and high baseload, as a percentage of Ontario demand. As the baseload:demand ratio moves about 90%, dumping, at depressed prices, begins. The higher this ratio, the lower the price and higher the net exports.
With the relationship established, we can now forecast what is most likely to occur with electricity market pricing in Ontario. The data for the following graph I developed for an earlier project, "The Real Costs of Wind Generation in Ontario." The historical figures demonstrate Ontario's move from net importer, prior to 2006, to large net exporter now. This trend coincides, as one should expect, with a decline in the market, HOEP, rate (from, approximately, $72.14 in 2005 to $32.60 thus far in 2011).
The number of hours where baseload exceeds Ontario demand is set to rise rapidly through 2014, when refurbishment work on nuclear reactors will temporarily slow the growth (in this model). However, the continued growth of wind that is currently foreseen will have a disproportionate impact on seasonal excess supply frequency. Because wind performs poorly in summer's heat (average capacity factors in the low teens), removing nuclear capacity will temporarily reduce hours where baseload exceeds demand – but the higher capacity factors of the shoulder seasons, and cooler periods, means that surplus baseload generation will escalate rapidly with increasing wind capacity even with a lower nuclear contribution in Ontario's supply mix. It will simply move from primarily a problem from May-October (the OEB's summer period, and also the period of most SBG periods in 2009), to a year round issue.
None of this information is new to those of us familiar with the data. My previous forecasting project concluded we are moving towards a supply mix where approximately a third of all wind output will need to be curtailed, dumped on export markets, or displace baseload hydro and/or nuclear, a conclusion reached earlier in the year by Aegent EnergyAdvisors.
The Market Surveillance Panel's Report urges the IESO to speed up the process designed to make, “wind generators dispatchable,” which is a solution to the market pricing, but a solution that is to pay suppliers without accepting supply from them. The IESO's process, SE-91, seems to have a worrisome focus, albeit one determined by government policy. Central forecasting may be a tool to determine how much to pay suppliers when their production is curtailed due to a surplus of supply – but there seems less interest in making the forecast accurate with the lead time, and frequency, required by the market design. Most of the 'stakeholders' are from the renewable energy industry, and as such wouldn't have any foundation of knowledge in the requirements of a healthy market.
That won't be much of an impediment in Ontario, where there isn't a healthy market, and none seems likely in the foreseeable future.