Late last year I worked up the likely amount of public money that would have to be thrown at the nuclear industry in order to successfully and safely decommission the 100 operational reactors and the now shut down ones. Unsurprisingly, the nuclear industry had been very optimistic in its estimates of decommissioning costs and timeframes, when the global empirical averages were trending to a billion USD and 100 years per reactor.
Recently I was asked by an Ontario journalist what I thought the likely situation in Ontario would be, and whether the decommissioning trusts were equally underfunded. I was unsurprised to find that Canada is in the same boat as the US, with highly optimistic schedule and cost projections which belie Canadian empirical experience with the CANDU reactor, and that the fund had nowhere near the money necessary for the job. Let’s run the numbers.
Ontario Power Generation (OPG) is the chunk of the provincial utility that was carved apart in the late 1990s by the Mike Harris Conservatives to handle generation alone. It operates 18 aging CANDU reactors across three sites: Bruce, Pickering, and Darlington.
OPG has a nuclear decommissioning fund of about $5 billion CAD or US$4 billion right now. If the experience of other countries on the actual cost of a billion USD per reactor and an actual timeline of decommissioning of a century holds true, and I see no reason why it doesn’t, that means that there is currently a $17.5 billion CAD gap in Ontario, in addition to the existing $19.3 billion CAD in debt still being serviced from their construction. When the government of the era split up the utility, it moved all of the debt off of the components and into general debt. One of the many appropriate and sensible things that the McGuinty Administration did in the 2000s, in addition to shutting down coal generation entirely, was to move the debt back into the utility and set about servicing it from utility bills.
Most of the reactors at Bruce Nuclear are aging out, with several over 40 years old and the remainder approaching 40. Darlington’s are around 30, so they have a bit of runway. Pickering’s reactors are going to be shut down in 2024 and 2025 and start decommissioning in 2028. While refurbishment could bridge Ontario’s for another 20 years in many cases, that’s expensive and typically won’t pass any economic viability assessment compared to alternatives.
The likelihood is that all reactors in Ontario will reach end of life by 2035, and be replaced by some combination of renewable energy and HVDC transmission from neighboring jurisdictions, with both Manitoba and Quebec having excellent, low-carbon hydroelectric to spare.
Does the empirical experience of shutting down CANDU reactors track to the roughly billion USD that’s seen for other reactors? According to the World Nuclear Association, no.
“The fourth unit is Gentilly 2, a more modern Candu 6 type, which was shut down at the end of 2012 after 30 years operation. It is being defuelled and the heavy water was to be treated over 18 months to mid-2014. A decommissioning licence was issued for 2016 to 2026 and the main part of the reactor will be closed up and left for 40 years to allow radioactivity to decay before demolition. All 27,000 fuel bundles are expected to be in dry storage (Macstor) by 2020. The decommissioning cost is put at C$ 1.8 billion over 50 years.”<
That translates to US$1.44 billion, so it would appear as if CANDUs are on the expensive side to decommission. If that holds true, Ontario’s gap is actually in the range of $27 billion CAD.
Nuclear decommissioning funding comes from reactors operating revenue. In the US, it’s 0.01 to 0.02 cents per kWh as a set aside. I wasn’t able to find the required set aside for Ontario’s fleet, but obviously they aren’t setting aside sufficient funds now, or have absurdly optimistic fund growth expectations. They only have a decade to set aside more money from operating reactors, and have only set aside $5 billion CAD after 50 years, so the most generous assumption is that they will set aside perhaps $7 billion CAD in the OPG fund by end of life of the reactors, and have a liability for decommissioning of $15.5 to $27 billion CAD. For the next step, let’s assume $20 billion CAD for the sake of round numbers.
Given the likelihood of all of Ontario’s reactors being off of the grid by 2035, with major decommissioning occurring every few years until then, the kWh generated by Ontario’s nuclear fleet from now through 2060 will be in the range of about 1000 TWh assuming there are no lengthy outages at any of the plants, which to be clear is an awful lot of low carbon electricity.
However, $20 billion is a big number too. It turns into about 19 cents per kWh if you only count electricity generated from today through end of life for the reactors. It’s obviously a lot lower if you calculated from beginning of the lifetime of the reactors. However you count it though, that’s only the unfunded Ontario liability, and it’s on top of subsidized security costs Canada and Ontario and municipalities bear, and it’s on top of the outstanding $19.3 billion in debt that has only been receiving servicing on the interest since the McGuinty government brought it back into the utility. It’s likely that the majority of that debt will be outstanding in 2035 still, as it has gone from $20 billion to $19.3 billion in the last 11 years, so expecting it to be gone by 2035 is not realistic.
So yes, Ontario’s nuclear program will be a fiscal burden on Ontarians to the tune of around $40 billion CAD which will be spent through roughly 2135, finally being paid off by the great-grandchildren of babies born in 2021.
Nuclear, the gift that keeps on giving.