Yesterday Statistics Canada’s daily news included Electricity supply and disposition, 2016.
Geoff Zochodne, a reporter at the National Post, gleaned this message from the release:
“Ontario exported more power to the U.S. last year than it has in a decade, and at a relatively low value.”
This initiated some e-mailing, which drew me into the data quagmire again, but also reflecting on my history reporting on exports. Instead of putting my thoughts into private e-mails, I thought I’d make them the content of this public blog post.
I’ll return to the newly posted Statistics Canada data later, but for now I’ll declare my bias as printed in the Financial Post early in 2016: “…StatsCan data is awful. It can’t be the basis for anything.” The recent release has mostly meaningful data, but some big errors mean it’s far from the best data to serve in analyzing Ontario’s exports – or anything else.
Some background on my involvement reporting on exports – if only to satisfy my sudden nostalgia.
Reporting on losses of exports is what got my blogging noticed back in January 2011. I’d started writing a couple of months earlier – to maintain skills in data analysis and, hopefully, develop some writing ability. January 1st, 2011 was warm (for winter) and it was windy. I wrote of records:
January 1st brought a new record that, if it became a trend, could lead to us wishing for the good old days where we paid suppliers $70 for a product to export at $35.
The IESO daily summary for January 1st, 2011, shows an average Hourly Ontario Energy price at a record low of negative $20.29/MWh. With exports averaging close to 3000MW each hour, we paid external markets about $1.5 million to take off our hands production we paid our suppliers about $5 million dollars for.
I created my modern version of a daily report for that date and it estimates the average cost of a generated unit of power on January 1st 2011 as $67/MWh, with exporters paid an average HOEP value of $22 to dump 72,325 megawatt-hours of supply: so today I’d peg losses on exports at $6.4 million that day. I like to think this story got enough attention to cause other stories, and by the end of January 2011 the Premier was addressing the topic in a manner I suggested, and still suggest, was less than truthful.
“I’m for having too much electricity .” – Dalton McGuinty
In the following months I received a phone call from Parker Gallant. The first published material from our relationship, Power dumping, was printed on July 20th, 2011.
Nothing that has occurred with export pricing since 2011, quantity, or losses has been more newsworthy that breathing to Parker and I.
1+1 equals 2.
That which must happen does.
The government continued to attempt to defend exports as profitable for many years, generally under the argument that all power purchased was necessary for Ontario’s system, and whatever revenue we received through selling to export markets was a bonus. A typical monthly news release from the Ministry of Energy, such as May 2012’s Electricity Exports Continue to Generate Revenue ( https://news.ontario.ca/mei/en/2012/05/electricity-exports-continue-to-generate-revenue.html ) would reference an IESO page that reported monthly exports, and monthly export revenue – that page, http://www.ieso.ca/imoweb/siteshared/imports_exports.asp , has long since been removed by the IESO.
After years of hammering away the point that revenues are not profits – those require subtracting costs from revenues – the IESO mostly hides the revenues on exports. Thus the excitement when Statistics Canada releases some flawed data that reveals exports have been rising as the average revenue per unit declines.
Let me demonstrate the most notable flaws in the Statistics Canada announced yesterday – specifically CANSIM 127-0008, Supply and disposition of electric power… I’ll compare the data to 2 other sources, but the differences in what is being measured need to be noted:
- The IESO has a page on their web site reporting imports and exports by jurisdiction: this measure is by grid intertie connection, so records only deal with jurisdictions that have a grid connection with Ontario.
- The National Energy Board (NEB) monthly reporting reports on the generation point and final sale location – so a purchase made on the Ontario market could sell to Maine after being wheeled through Quebec
- Statistics Canada claims it uses survey data
Here’s a graph of how the 3 annual data sets compare from 2005-2016:
You don’t need to be a statistician to see there’s a big problem with the CANSIM 127-0008 data for 2013 and 2014. In the other years it’s very similar to either the NEB data from 2008-2012, and in 2005: it’s close to the IESO data in other years (2006, 2007 and 2015) – but for 2013 and 2014 that’s simply bad data.
The cost side provides a little bigger challenge to get to an IESO figure for comparison. I think the best data I have on what the Ontario system receives for exports was revealed through a response to a freedom-of-information act request, but that figure includes exports through interties to other provinces as well as states. I have recklessly assumed exports are similarly priced and scaled down the revenues to the U.S. share of exports.
Here the new CANSIM data for revenue, from Statistics Canada, closely match the IESO data until 2013, where luckily it diverges in a very similar way to the divergence in the electricity volume’s diverged.
A point of interest from the revenue charting is the significant difference between the “IESO FOI-base estimate” and the National Energy Board figures. While the IESO figure is speculative, the difference has to exist for traders to be active: the IESO figure reflects traders’ purchase pricing from the Ontario market while the NEB measures revenues from the re-sale into foreign markets.
Throwing together the relatively flat revenue and escalating export volumes, we find declining average rates ($/MWh) paid for Ontario electricity. I’ve added to the confusion by including the average weighted value of all exports (including to other provinces) at the Hourly Ontario Energy Price (HOEP), which is the only figure I can use for hourly analysis. I wanted to show this figure is a little lower than what the IESO reported as actual received in the freedom-of-information requests. That difference can be explained, but I just want to demonstrate here the Ontario electricity system’s export revenue is usually a little higher than I can estimate with consistently available data, but not so high as the NEB data would indicate (as it includes the arbitrage in the trade). The new Statistics Canada data does indicate a rate in the ranges of the other data sources, but after 2012 it’s very wonky indeed.
The IESO reported the average wholesale rate for electricity in 2005 was $78.93 in 2005, but that was an exceptional year for pricing which dropped to $47.80 in 2006.
In 2006 Ontario consumers paid about the same for Ontario power as did American consumers.
In 2016 Ontario consumers paid, on average, $113/MWh. American consumers of the same electricity paid a fraction of that at between $20 and $30 per megawatt-hour.
We can’t be certain of the current rate for exports, but it’s certainly a lot less than Ontarians pay suppliers to produce the exported electricity.