Tuesday, January 7, 2020

Review of Annual Ontario Electricity Data

It's that time of year. This will be the tenth time I've produced a post on year-end electricity data. In my first such post I quoted the IESO's release on data for 2010:
"The cost of power in 2010 was 6.52 cents per kilowatt hour (kWh), as compared to 6.22 cents/kWh in 2009. This cost includes the average weighted wholesale market price of 3.79 cents/kWh and the average Global Adjustment of 2.73 cents/kWh (preliminary)."
It's a little higher than that this year. A similar sentence for 2019:
The [class B] cost of power in 2019 was 12.63 cents per kilowatt hour (kWh), as compared to 11.51 cents/kWh in 2018. This cost includes the average weighted wholesale market price of 1.8 cents/kWh and the average Global Adjustment of 10.8 cents/kWh (preliminary).
The prices aren't strictly comparable for two reasons, but for most consumers the difference will still be significant.
The numbers are nominal, but there was little inflation in Ontario over the decade (approx. 18%, and 1.85% in 2019) so the real "cost of power increase" for most consumers was still 65% over the decade, and 7.8% last year, which means last year was worse than average!
The "class B" distinction is necessary as two - or three - distinct consumer classes were created over the past 10 years.

I'll look at three distinct areas of supply: the reported generation figures on the IESO-controlled grid (ICG), distributed generation, and curtailed generation (which is supply Ontarians will pay for but was not accepted onto the ICG). I'll look at costs by the fuel, or supply type, as the IESO reports for generation. I'll look at "the cost of power" by consumer groupings, and provide an average cost of power. Most of these figures are estimated, and even figures produced by the IESO will often differ from one another due to minor differences in the origin of the data.
I will not generally try to reconcile my estimates to figures reported by the IESO due to limiting my time - and not agreeing with the IESO on discrepancies I've found in the past. Please interpret numbers as illustrative, and not gospel. This should be particularly obvious for cost break-downs which are not available anywhere else: most contracts are private and rates are estimated as best I can. I will briefly discuss the IESO's monthly global adjustment component cost file, which is one place you could find some confidence in the quality of my estimates, and/or their accounting.

I'll try to maintain a focus on providing a basis for analyzing opportunities to reducing the cost of electricity in Ontario in preparation on a future post on impediments to cutting costs.

A first metric for those who see little opportunity for cost cuts:

The average price paid to a supplier for a single megawatt-hour (MWh) was less than $94 (or 9.4 cents per kilowatt-hour), 35% below the average rate paid by Class B ratepayers. 

Now for the many numbers needed to make that conclusion.

View these estimates with Google Sheets

The format of these numbers is one I've revised slightly since writing on it a couple of years ago - by deleting the section calculating rates which I feel can be harmful to understanding the value of each type of generation. The format originates in the IESO's response to a freedom of information (FOI) request back in 2017 . Note I have not adjusted my estimates to match the FOI response's figures for 2007-2015, but I did use the figures to guide my rate estimates for individual generators (which are the basis for the summary figures shown in the table).

The "TWh supplied" estimate includes both the reported generation on the IESO-controlled grid (ICG) and supply embedded within the grids of local distribution companies (LDC's). It does not estimate behind-the-meter/fence self-generation as I have no way to do so. There was a report in 2018 that 1100 megawatts of such capacity exists, so the annual generation need not be trivial and could have a growing impact in lowering demand for electricity from the ICG.

Subtracting exports from all supply should yield an estimate of consumption in Ontario. It is down in 2019 by about the same amount as the IESO's "Ontario Demand". The latter figure excludes supply from embedded generators, which I estimate changed little in 2019. I assume the 1.7% drop is due to weather, but am not equipped to calculate that.

The "Cost $Ms" section includes all costs which can be categorized as either energy (output), capacity (paid for being available), curtailment or conservation. The differences will be clarified in the sections discussing the performance of separate generation types.

The biggest dollar figure in the cost components is, in my opinion, the largest story of the year: $7.5 billion spent on nuclear supply is an increase of $1 billion from 2018. All other supply costs were up only $275 million.


Nuclear output was essentially flat in 2019, while both nuclear suppliers realized significantly higher rates for their supply. I estimate nuclear supplied 55.5% of all electricity at 48.4% of total costs.

The Bruce Power rate hike occurred under the terms of the contract/agreement for the refurbishment of 6 reactors. One reason "subsidy" became a word I try to avoid is the complexity of arrangements. In the old days of a single public integrated utility the costs for a project weren't to enter the rate base until the project was completed and entered service. Lower rates can be now be given for projects by recovering costs as they occur, prior to the project entering commercial operation. 

I estimate the average cost of supply from Bruce Power rose to $77.64/MWh in 2019 (up from ~$67.24/MWh in 2018), but that's still well below the average $94/MWh I noted above, and far below the $100/MWh I estimate all other generation cost in 2019. 
The rate change for Bruce Power's nuclear production increased supply costs by approximately $323 million.

I was hoping to keep the discussion on costs away from the global adjustment - which is the difference between what is paid for supply (generally either due to contracts or regulated rates), and what is recovered through the system operator's (IESO) attempt at a market rate (HOEP), but need to mention it briefly here. The IESO has been reporting a breakdown of the global adjustment components (in MS Excel), and just broke up a category that included both non-public gas and nuclear (at this time they've bungled the data for 2015-2017, but the last two years looks right). The relevant, and exciting change in the IESO's new global adjustment breakdown, is the new "Nuclear (non-OPG)" category which reports the global adjustment for just Bruce Power. Now that the IESO has broken the taboo about revealing costs of a single supplier they should be close to revealing the global adjustment charges for other specific suppliers - like Pattern Energy's generators being sold to the Canada Pension Plan, and/or the OPG/Lower Mattagami Energy Limited Partnership's generators.

This IESO's global adjustment reporting will eventually show, by mid-February, the global adjustment charge at Bruce grew about $635 million in 2019 - which looks worse than the actual cost increase because the HOEP value of their production fell by ~$313 million. The actual change is the supply cost of production from Bruce Power was therefore up about $322 million. (and that confusing distinction is why I wanted to avoid discussing the global adjustment!)

I estimate the increased cost for the production from OPG's nuclear facilities grew $688 million in 2019. Not only did the rate grow (to nearly $90/MWh), but output grew about 2.6 TWh from 2018 due entirely to the oldest units, at the Pickering Nuclear Generating Station, producing more power in 2019 than they have in 25 years. Accolades to the operators. But the opposite for the fiscal people at OPG.

I first wrote, alarmingly, on OPG's 2017-2021 rate application in June of 2016. The regulator (OEB - Ontario Energy Board) finally allowed rates from that application to first be recovered in March of 2018. The period without a rate set saw OEB's payment for its nuclear output drop from $72.30/MWh to $59.29/MWh, and a portion of today's higher rates is due to rate riders added to pay OPG for the lower rates of 2017.

When the OEB did get around to allowing new rates at OPG it did so based on the annual revenue requirements at OPG, and projected output. This year's very large increase in payments to OPG for nuclear is both because of higher rates in 2019 compensating for lower rates during the deliberation on the last rate application, and 2019 production at OPG's nuclear unit exceeding the production base from that application by 4.5 TWh (over 11%).

OPG executives may be pleased with the nuclear units performance the past couple of years, with the approval of rate increases based on much lower output than they've realized. The balance sheet must be strengthened as there's been a buying spree: acquiring Eagle Creek's US hydro sites last year served as a warm-up to a $1.12 billion purchase of Cube Hydro (sites in US),  and this year's purchases of all or part of four natural gas combined-cycled generating stations in Ontario.

I concluded an article on this topic in July with, "Until [conditions are met]...I won’t advocate for the refurbishment of the final two units at Darlington." I've seen nothing to indicate OPG will address nuclear rates, and some more evidence its interest in nuclear has waned. Time was of the essence in the Darlington refurbishment project, but when the first refurbishment (Unit 2) fell behind schedule the second (Unit 3) was delayed. It's not unimportant to note that the government approved the Unit 3's refurbishment 22 months ago. Unit 1 was scheduled to enter refurbishment mid-2021, but 18 months out there is no confirmation that will happen.

After fighting off nuclear opponents for half a century, OPG's next request for a rate hike now appear to be the industry's biggest threat.


Like nuclear, hydro output in 2019 was essentially unchanged from 2018's production. I estimate hydro supplied 22.7% of all electricity at 13.4% of total costs.

It's likely production could have been higher (due to water conditions), and OPG's third quarter results confirm more "production foregone" due to surplus baseload generation (2.9TWh Q3-YTD, up from 2.4 TWh in 2018).

Earlier I alluded to showing the global adjustment for a private hydro generator. I estimate private hydro contracts receive a rate 2.5 times the size of the rate for the public generator's hydro output. Private hydro contracts should not be confused as being affordable: it's the legacy OPG sites' regulated rates that push down the average.

I do have a routine estimating hydro "curtailment". Although I'm sure it's far less accurate than my estimates for nuclear, wind and solar, the overall trends I capture do match the trends indicated in OPG's quarterly results. I don't attach costs to hydro curtailment as those are captured through rate riders in future years - but only for OPG. I suspect this is not the case for contracted generation and that the IESO has been negligent in omitting from its annual summaries the amount of curtailed hydro it pays for (and then collects from ratepayers).


Like nuclear and hydro, gas output in 2019 was essentially unchanged from 2018's production. I estimate gas supplied 6.5% of all electricity at 9.3% of total costs - but it would be unfair compare gas-fired generators to others based on the cost/MWh.

Most generators fueled by gas are now on capacity contracts with the system operator (IESO), with most of the contracts written by the Ontario Power Authority (OPA) before it was merged into the IESO. These contractors receive a net revenue requirement in order to be capable of generating when other contracted/regulated supply (nuclear, hydro, wind, solar, etc.) can't meet demand.

This was not the case prior to the current market design. The last time governments looked for alternatives to nuclear non-utility generators (NUGs) were contracted, most of which were fueled by natural gas, and most of those were paid volumetrically on a must-take basis, meaning the system had to buy all that was produced. There are still a handful of contracts that are similar, but most NUG contracts have expired (or been bought out). Most current gas facilities are paid primarily for their availability, and when they produce the incremental output is inexpensive (as I noted in a 2019 blog post).

While the output from natural gas fueled generators in 2019 was considerably higher than it was in 2017 ( a year of small demand and huge excess supply), it's the second lowest year for output since the inception of the IESO market in 2002 - and perhaps for quite some time before that. Most of the non-utility generator (NUG) contracts were from the Liberal and NDP governments of the late 1980's and early 1990s, and most gas generation was from the now-expiring NUG contracts.

I noted earlier the publicly owned generation, OPG, was on a buying spree. It included the purchase of the remaining half of the Brighton Beach Generating Station in Ontario (combined-cycle gas), and then:
In July 2019, OPG, under a new subsidiary, entered into a purchase and sale agreement with affiliates of TC Energy Ltd. to acquire a portfolio of combined-cycle natural gas-fired plants in Ontario for $2.87 billion, subject to customary working capital and other adjustments. The acquired portfolio includes the 900 MW Napanee GS, the 683 MW Halton Hills GS, and the remaining 50 percent interest in the 550 MW Portlands Energy Centre.
The biggest prize here is probably the "Napanee GS" as it's the largest unit, contracted at the highest net revenue requirement/guarantte, with the longest time left on a contract. The value was created in the strangest way: Napanee is the rebranded Oakville Generating Station, which began as a joint proposal by OPG and TransCanada. The two companies had worked well on the Portlands Energy Centre, but by the time the Oakville project was contracted the government of the day had decided OPG shouldn't be involved in any new gas-fired generation and TransCanada was on its own. Without OPG's competency in working with local communities to get projects built the opposition to the project grew to the point the government withdrew approval for the location during an election campaign (one that featured all parties promising to do the same). Gas plant scandal documentation and Auditor General reports chronicle the price rising as the government searched for a location to substitute for the one needed in Toronto's western suburbs and appease TransCanada for reneging on the initial contract. Now, after a couple of years ramping up nuclear rates, OPG has bought the contract for the project it was banned from participating in, after that project was relocated to a site OPG owned. It paid a price multiple times higher than it could have built and operated the facility for independently.


Wind output didn't increase a lot in 2019, but it added more megawatt-hours than any other source in the IESO's generator reporting. I estimate wind supplied 7.8% of all electricity at 12.3% of total costs - but as with gas there is a factor to be considered before simply dividing cost by generation to get an average rate.

Despite outgrowing traditional generation sources wind grew more slowly in 2019 than in any year since it appeared in records in 2006. 

During the final days of 2019 output appeared in reporting for the Romney facility. Earlier in December the government had revoked the approval of the Nation Rise project, which made Romney the sole remaining contracted project not in commercial operation. With Romney now operating, the pipeline is now empty.

Earlier in the year - but not much earlier - the Henvey Inlet/Greg Rickford industrial wind facility came online. I won't beat this drum for too long here except to note the IESO Contract List data shows the facility entered commercial operation over 8.5 years after the date it was contracted under a feed-in-Tariff program that paid premium rates due to a claim that time was of the essence and commercial operation was to be reached within 3 years. That time to commercial operation was easily a record for Ontario wind contracts. The date for commercial operation also exceeded the heavily modified Milestone Commercial Operation Date by 561 days - which wasn't a record but did exceed the 18 months specified in the standard feed-in tariff contract's reasons-for-termination section. The results of Ontario's most expensive wind contract entering service later than all but one other facility is it utilizes the latest, most efficient turbines. The lack of action on the estimated $150/MWh contract made it an appealing purchase for the Canada Pension Plan.

As with the Oakville/Napanee generating station, wind generators in Ontario sell for more than it costs to build new ones because the contracts are so rich.

Wind is curtailed more frequently than any source except hydro. It's unclear how much curtailment is due to surplus baseload generation (the system simply having too much supply), and how much is instigated by the system operator to add flexibility, which in the past was provided by coal and gas-fired generators. In recent years curtailment is occurring during the peak solar hours of the afternoon. 

Image captured from Power BI wind report: 2019 hourly (EPT) average capacity factor (CF), estimated percentage of potential generation curtailed, and adjusted CF (including all potential generation).

The cost of that mid-afternoon wind curtailment could be considered a cost of solar. Curtailment in general might be considered a system cost without assigning the expense to any particular source, although that would ignore the reality that curtailment generally occurs when it's windy.


Solar had the highest percentage growth in 2019, at least in the IESO-reported generator data.
I estimate solar supplied 2.4% of all electricity at 11.3% of total costs.

Most solar is not in most reporting as the majority of solar capacity is installed in distribution networks.  2019 was anomalous in that more capacity was added on the ICG (reported sites) than on local distribution networks. Reviewing my figures now I'm probably a little optimistic in my estimate on distributed solar production.

While not as dry as the wind project pipeline, there's very little left in solar's pipe. There were still a smattering of projects not in commercial operation as of September 30th, 2019, but only about 35 megawatts of capacity (not inclusive of micro-FIT).


Imports declined to 4.1% of supply in 2019. I estimate they were only 1% of costs, but it's likely a little higher than that as some imports, from Quebec, are contracted at an undisclosed rate (rumoured to be around $100 million for 2 TWh each year).


I'll finish the discussion on supply by referencing an article I wrote in 2019: Available low-cost electricity not utilized in Ontario. With demand down and contracted, or regulated, supply up slightly, rates were not going to fall. 

Higher consumption would drop average rates as it would reduce curtailment while increasing imports and/or supply from natural gas-fueled generators.

Consumer Segment Pricing

In the first section of this review I noted, 
The average price paid to a supplier for a single megawatt-hour (MWh) of supply was less than $94 (or 9.4 cents per kilowatt-hour), 35% below the average rate paid by Class B ratepayers. 35% below the average rate paid by Class B ratepayers. 
I'll quickly connect the review of supply with consumer costs by connecting the $94/MWh paid, on average, to suppliers, and the $126/MWh average I expect to be charged to the basic Ontario ratepayer.

Not all purchased supply is delivered: some is curtailed either due to a current surplus or to provide flexibility to the system, and some is lost in transmission. Some costs that reach end consumers are for conservation, and some are for storage. When I look at rates from a consumer perspective - adding up the value at the so-called market's rate (HOEP), and adding the global adjustment pool - the average rate paid per megawatt-hour for all consumer groups in 2019 was a little over $100.

Some consumer groups paid a lot less than the average, so others paid a lot more.

Graphic is explained in Trends in Ontario Electricity rates by Consumer Segments
There are primarily 3 consumer segments: exporters, Class A, and Class B. For a period Regulated Price Plan (RPP) consumers were separated from "Class B"  by the original [un]Fair Hydro Plan scheme. I won't explore that further here as going forward the RPP rates will return to being a forecast of Class B rates (periodically adjusted for variance to realized class B rates).

The global adjustment is charged to consumers monthly to recover the supply costs not recovered through sales in Ontario's so-called hybrid market. Exporters aren't charged the global adjustment, but they don't only pay the market rate either. A market control price (MCP) at interties sets the value of exports (not necessarily directly - but that detail is not relevant to this review). The difference between the MCP and the HOEP, which I'll call "congestion rent" is not used to reduce the global adjustment, but does benefit consumers through reducing Wholesale Market Service Charges (WMSC).

The HOEP value of exports was over $1.5 billion less than the value of exports at the average cost of supply in 2019, so that amount needed to be recovered from within Ontario via the global adjustment mechanism. This added about $11/MWh to the average Ontario rate.

Global adjustment charges do not get allocated proportionally by usage across the two classes of consumer in Ontario: A and B. Class A (larger) consumers can avoid global adjustment charges by reducing consumption during the five highest daily peak demand hours of a 12-month period under a program called the Industrial Conservation Initiative (ICI). Theoretically this could benefit all Ontario consumers by reducing the need for capacity to meet higher demand peaks.

In 2019 the ICI transferred over $1.4 billion from Class A consumers to Class B consumers, which adds $14/MWh to Class B consumers' bills. That's a simple number to quantify: less easy to quantify is the benefit to the system of avoided peaking capacity. There is a very good indicator of whether the ICI provides good value though: it's up in the review section on natural gas where capacity charges of about $900 million a year are shown. Those capacity charges pay for over 8,000 MW of capacity. Average Class A consumption is 2019 was about 4,600 MW, so even if Class A consumers shut down completely during peak hours class B consumers would still be paying 55% more for 45% less than the value provided by existing natural gas generators.

In summary: an average supply cost of $94 is a base upon which a couple of dollars are added for curtailment, another couple for line losses, a few more for conservation, $11 more as penance for buying too much and dumping it on export markets, and another $14 to switch costs from large consumers to small ones.

In 2019 the government subsidized small/voting electricity consumers at a cost to the provincial treasury of over $3 billion - roughly equal to the addition to those voters' bills from dumping exports and the cost shifting of the ICI.


No comments:

Post a Comment