November's Record electricity Rates: the Global Adjustment/Ontario Roulette story

Sources inform us Ontario's common ratepayers will pay more for November's electricity than they have paid in any month previously.
That may be true for some but it's probably not true for all.

The Toronto Star's John Spears courageously tackles the "murky fee" that is the global adjustment (GA), but that charge is primarily the difference between the cost of supply, almost all of which is contracted, and the share of those costs recovered by sales at the Hourly Ontario Energy Price (HOEP).  The primary reason for record rates are the contracts - not the GA; the "actual" GA for November is higher than it's ever been largely because the weighted HOEP average for the month is the 2nd lowest it's ever been.

The Independent Electricity System Operator (IESO) reported November's final "global adjustment" figures at a record $847 million dollars, resulting in a record "actual rate" of $78.55 (class B).  Combined with a weighted average Hourly Ontario Energy Price (HOEP) of $16.08 (the 2nd lowest ever), the wholesale market commodity charge will be reported as a record $94.63.

My estimated composition of November costs (view as webpage)

My estimate of the global adjustment did indicate a record, but not as high an amount as the IESO now reports as "actual."  That's not surprising as there are a lot of murky areas in estimating the total value of the electricity sector supply, and the total demand to allocate it to. [1]

A quick overview of the estimates in the table:

1. Estimated generation, market value and contract cost 
2. Estimated embedded generation (contracted, but not directly on the IESO grid)
3. Estimated curtailment levels 
4. Estimated Global adjustment as difference between supply costs and supply value at HOEP
5. Balance the estimate to much higher ($72 million) actual global adjustment reported by the IESO
6. Reduce demand by estimated line loss
7. Remove exports from total consumption and market value (valued at HOEP)
8. Estimate cost to class B customers of lower Class A charges

Confusing, and with a lot of assumptions, but far better than no estimate at all.

The price hit a record in November by a little bit - and that was determined by record production from Ontario's industrial wind turbines and strong output from nuclear generators which, during the first 2/3rds of the month, meant very low pricing (~$10/MWh - the month's 111 hours of negatively pricing hours is the 2nd highest monthly total since market inception); the end result being that an average cost per megawatt-hour of $81.52 gets bloated by ~$10 to cover the export sales that collected only ~$13.64/MWh.

There's a lot covered there, but it's all been covered on this blog before; we know the global adjustment mechanisms alters the distribution of costs across market participant groups.

We know the "Class A" global adjustment scheme was created to do exactly that.

Operationally it appears the global adjustment may be applied in a manner that distributes costs inequitably based on billing cycle dates.

In the estimates above, adjusting to the IESO's Actual $847 million global adjustment is a larger difference than I've experienced most months [2], and I suspect that is because October's $634.3 million "Actual" global adjustment was lower than it should have been.
The premise that there is a "true up" between months is interesting in terms of the premise behind the global adjustment; that the full cost of power be paid by the customers that consume it.

The idea that charges can be incurred in one period and allocated in another probably might not be unsettling, but how about the related idea that charges for electricity consumed at the same time will be priced differently between customers of the same class, based on their billing date?

This from the IESO site:

What is the monthly rate for Global Adjustment?
There are three rates, each available at different times of the month. Each is set to recover the total monthly costs of the global adjustment. Depending on when in the month you are billed, your LDC may use any one of the three rates. The 1st estimate is posted by the first business day of the billing month, the 2nd estimate is posted by the final business day of the billing month and the actual global adjustment is posted on the tenth business day after month end. Monthly estimated rate information for 2011 and 2012 is available here.

I interpret this as telling me that the November global adjusment charges, paid by similar customers, can be:

• $62.28 for the initial portion of the month (for any billing period that doesn't match a calendar month)
• $84.89 for the November section of a billing period ending in December prior to the tenth business day (or for a billing period that does match the calendar month)
• $78.55 for the latter portion of November for billing periods after the 10th business day of December

The global adjustment rate for December is currently set (1st estimate) at $76.07/MWh.

I see absolutely zero possibility that the "actual" rate will end that high.

None.

This seems like a lottery system for collecting electricity payments.

Table from IESO website

As with all costs contained within the mysterious bubble that is the global adjustment, there's no transparency on how costs are reconciled.   True-ups surely occur, and there's some evidence that is done by moving costs between months with the 1st and 2nd estimates:

• in June the 2nd estimate was $15/MWh more than the first, so ...
• in July the 1st estimate was set high, and the 2nd estimate was $28/MWh lower, so ...
• in August the 1st estimate was set low, and the 2nd estimate was ~$29/MWh higher, so ...
• in September the 1st estimate was set high, and the 2nd estimate was ~$24/MWh lower

It appears that because of the billing cycles the "true-up" is in the 1st estimate of the next month.

Currently, a customer who escaped November's record rates by getting the benefit of the $62.28 rate for most of the month will be looking at a December Global Adjustment rate of $78.55 - even though that's not a sane estimate for what would actually be reflective of costs incurred in December.

The Global Adjustment mechanism is the subject of a stakeholder initiative at the IESO.

The first question in reviewing changes to any mechanism might be "evolution or revolution?"

Revolution.




ENDNOTES

[1] The IESO reports on supply from generators on the IESO grid; most, over 10MW capacity, show on the "Generators and Output Capability Report", and the total, including the small generators omitted from that report,are included in the figure the IESO reports as "Ontario Demand" - because the IESO does not report actual demand, but the total of all generation on it's grid.  On the pictured spreadsheet, the difference between "Ontario demand" and the sum of generation on the Generators report is indicated as "self-scheduling"
One challenge is the generation contracted but embedded with generators and not on the IESO's grid.  That generation currently includes all solar generation in the province (and I've estimated 10% more wind capacity than reported by the IESO).  There's likely a number of older generators in that category too.  

Regardless, we don't have a good grasp of either what supply actually is, or what demand actually is. 

[2]  The estimates included here build on the foundation of the work I do reporting "supply costs" which appears on this page (and I update twice a month).  New to the analysis on this page is reducing the consumption/denominator by 3.2% (as a line loss adjustment).  

YTD, my estimates are $482 million shy of the IESO's actuals; of that I believe $225 million is charging for the costs involved with relocating the Oakville Generating Station away from Oakville.

The remaining variance works out to 3.6% - while I can't explain it (or there'd be no variance!), it seems to me likely due to embedded generation.  

Without the Ontario Power Authority providing details on their payments to embedded generators, it's not possible to get particularly accurate either on the cost of supply components, or the actual metered demand to distribute those costs across.