The High Costs of Ontario's very provincial electricity debacle

Ontario's latest electricity sector fiasco is being credited as a major factor in Premier Dalton McGuinty's announcement that he is proroguing the legislature (done), and resigning (not done).  There are some important lessons the global community can learn from the poor example Ontario has provided.

Many regions of the world are struggling with an electricity sector design that can maintain sufficient reliable capacity to meet demand at all times while encouraging intermittent generation from renewable sources that may, or may not, be present when demand is.  The debate on how best to accomplish that in other jurisdictions exists - in Ontario it does not.
While globally jurisdictions weigh the pros and cons of an emergency reserve option, (such as in the Nord pool market), the option of capacity markets (used fairly widely in the United States), or more rigorous attempts at allowing a market to co-ordinate supply and demand (such as in Texas), Ontario displayed it's family compact heritage in providing secretive private contracts that guarantee a monthly net revenue to the lucky recipients of the contracts.
The Net Revenue Requirement (NRR) is a form of capacity payment.

Ontarians should get a full accounting of the cost of moving the plant, but that's not the big cost of the energy policies.

Other jurisdictions should review the tale of a province which embraced capacity payments to find, predictably, that the power generation industry stopped being about the efficient generation of power and became entirely focused on producing contracts that are profitable regardless of whether or not power is produced.

---

The Plant That is Relocating a Premier

This week, our government announced we are relocating a gas plant from Oakville to eastern Ontario. The total cost of the relocation is $40 million. 

-Premier Dalton McGuinty, September 24th

The cost of just the move is best estimated at almost 20 times that amount:
The most credible estimate of the cost of relocating the plant is $733 million, as presented by Bruce Sharp.

The government continues to be under well deserved criticism for the costs of relocating the natural gas generation station - both in terms of the actual expense and in terms of hiding the expense from the legislature.  I'll quickly review the history, and note some various figures on the cost being presented, before sharing estimates of the ongoing costs of the electricity plan that has natural gas-fired generation facilities with Net Revenue Requirement (NRR) contracts as a core component.

OPA Graphic: 4 qualified bid locations

• On April Fool's Day, 2009, the Ontario Power Authority (OPA) announced four firms had qualified "to bid on a new natural gas power plant to serve the electricity needs of the rapidly growing southwestern GTA area."
• On September 29th, 2009, the OPA announced, "it will sign a contract with TransCanada Corporation to design, build and operate a 900 megawatt (MW) electricity generating station in Oakville to provide a new, cleaner source of electricity for the growing southwest Greater Toronto Area. This new natural gas power plant will maintain local supply reliability and replace the coal-fired Lakeview generating station."  (note: the original bidder was Portlands Energy L.P. - which is 50/50 Transcanada and Ontario Power Generation partnership).  
• On October 7, 2010, the government announced the Oakville Power Plant would not be moving forward with the explanation, "When the need for this plant was first identified four years ago, there were higher demand projections for electricity in the area."
• On September 24th, 2012, the Minister of Energy announced "an agreement has been reached between the Ontario Power Authority and TransCanada Energy to relocate the proposed 900-megawatt natural gas plant originally planned for Oakville to lands at Ontario Power Generation’s Lennox Generating Station site near Bath, in eastern Ontario’s Lennox and Addington County."

To recap: a plant was sought for one of the few areas of the province still growing demand (demand in Ontario peaked in 2005).  The plant was cancelled with the explanation that demand in that area is now not growing very much.

Your first loss is your best loss

I think many in Oakville would be familiar with this investment saying.  

The relevant, fair, question to ask is what is the cost of this relocation, including the annual additions to the cost of electricity in Ontario, versus the cost of paying the penalty and cancelling the plant.  The cancellation liability rumoured at the time of the cancellation, in 2010, was $1 billion (here).

The plant is now reportedly guaranteed a Net Revenue of $15,200/MWmonth.  For a 900MW plant that equates to $164 million each year, or "$3.28-billion over the 20-year life of the contract." [here]
If the plant is not needed, the cost of relocation and ongoing subsidizing of the plant is far worse than the cost of cancellation.  The argument provided, that the plant is no longer needed to service growth in the southwestern GTA, is supportive of an argument that the plant is not needed - not that the plant should be located elsewhere.

The announcement on the Oakville plant relocation costs emphasized the NRR for the new project had been negotiated down to $15,200/MWmonth (~$2000 more than the fleet average of two months earlier) - down from $17,277 under a 2009 contract that was was negotiated at the height of the recession as Ontario demand was collapsing.  The announcement should have raised some eyebrows in that it followed, by only two months, communication that the average Net Revenue Requirement (NRR) for Ontario's gas fleet was $13,187/MWmonth, and the most recently contracted simply cycle gas turbine (SCGT) supply was contracted at $12400/MWmonth.  At $2000/MWmonth above the average NRR the TransCanada Oakville contract is over $20 million a year richer, and at $2800/MWmonth above the recent SCGT NRR, the contract is $30 million a year richer.

A more desired statistic, to most, would probably what the cost is per MWh.
This number was not released, and it's not particularly clear if the government is even capable of estimating it.

Levelized Unit Energy Costs - Scenarios

For the most basic estimation of an average price per MWh for the output from the plant, there are two variables in addition to the capital and non-fuel operational costs (which are conveniently included in the NRR): the capacity factor (CF -simply the percentage of theoretical maximum annual generation that is produced), and the cost of the fuel (which we assume to be 7.5 times the price per MMBTU of natural gas).

I have highlighted estimates at a variety of CF's which are calculated from forecasts for generation capacities in Ontario's Long Term Energy Plan (LTEP), and forecasts for production from Clearsky Advisors (page 38 here), the Pembina Institute, the US Energy Information Administration (page 48 here), and a US EIA levelized cost figure shown by the Ontario Power Authority in their system planning (slide 39 here). [1]

Production from Ontario's existing fleet of combined-cycle gas turbine (CCGT) generators with Net Revenue requirements over the previous 12 months has been at a capacity factor of approximately 35%, with gas pricing near historic lows - but that can't be interpreted as meaning the price to Ontarians, inclusive of the Net Revenue Requirement, is only in the $82/MWh, because much of that generation was sold in export markets at far lower rates ($25-30/MWh).  Attributing all net exports that could be attributed to the output from these generators (Portlands, St. Clair, Goreway, Halton Hills and Greenfied), almost half of all the generation can be considered for export. [2] 

The reason for the export is the net revenue requirement, which allows generators to bid into the market at any price that covers the incremental cost of generation (the fuel itself). Ontario has been a net exporter since 2005, which was the year demand peaked. The exports are broadly assumed to be in off-peak hours due to Ontario's large baseload generation capacity, and the accompanying surplus baseload generation (SBG) issue, but net exports have actually grown the most in the higher demand hours.

The range of forecast generation from natural gas is likely due to models that exclude exports and those that include exports despite the likelihood exports will remain close to the cost of fuel, while the capital costs will be paid entirely in Ontario through the net revenue requirement (via the global adjustment mechanism). The real capacity factor to utilize in estimating levelized unit costs for Ontarians, over the past 12 months, is ~20% (filtering out the exports). In the past year that means Ontarians have paid about 13 cents/kWh for the output of the NRR generators, while export customers have paid under 3.

The recovery of the cost of capacity payments is problematic for free and fair trade of the electricity commodity - transferring the generator's capital cost recovery to only the domestic ratepayer.   With the additional 900MW added to the total contracted capacity of NRR generators, there will now be approximately 5,830MW of contracted capacity at an average $13,457/MWmonth, for a total of about $940 million each year.  This is a very significant distortion to Ontario's market which should spill over into adjacent markets.

The Net Revenue Requirement does make one truth particularly clear - to get cheaper unit output, generating capacity  needs to be more productive.

Cost Will Continue To Increase If Ontario's Long-Term Energy Plan Is Implemented

Examining Ontario's planned supply mix it is apparent that a planned 10700MW of renewable capacity (8000MW wind and 2700MW solar) is not expected to have a significant capacity value - which means that Ontario is procuring the renewable supply and, concurrently, natural gas generation to ensure supply that can meet demand.

Currently solar capacity is slightly over 500MW, with wind around 2000MW.   If these figures grow the capacity factor of the gas plants will continue to drop.  

Utilizing a model I built based on a study in the UK (background here), it's possible to produce some figures to demonstrate the massive expense involved in providing feed-in tariff contracts to 'renewable' sources, where all supply must be taken, coupled with granting net revenue guarantees to dispatchable generation to run only as necessary.

This model is imperfect, but it does illustrate the cost of producing with renewables while natural gas-fired generators are idle. [3]

$3 billion a year.

That's a significant figure in a market that is currently about $10 billion a year.

It's notable that even in the scenario where no renewables are present, the natural gas facilities are still below the productivity level of CCGT facilities in the United States, where lower gas prices have seen a switch from coal to gas-fired generation that has brought fleet capacity factors up to where, I'd estimate, they are now over 50%.  Ontario is making it's natural gas generators a source of inflation where elsewhere on the continent they are acting to moderate pricing.

Assumptions stated in this post allow for estimating costs, but they are not entirely realistic.   The most significant issue is that the higher the production of intermittent generation on the grid, the greater the need to have traditional generators ready to adjust along with changes in output.  There are significant implication of that which include the wisdom of using CCGT plants at all.  Coal units are proving to be more complimentary as they provide far more peaking depth.  Gas turbine developers are attempting to deal with that challenge: only days after the Ontario government agreed to purchase the turbines TransCanada had hurriedly ordered for an Oakville facility that was unlikely to ever exist, GE started marketing a turbine with much greater operating range and ramping speed.

TransCanada will be guaranteed healthy profits to build a 2010 Oakville facility in 2017 Bath.
The losses from that decision will continue for 20 years.

Notes

[1] Most figures in this post are contained in this spreadsheet, including a more complete version of the $/MWh spreadsheet embedded in this post.
[2] Net Exports are attributed to the generation for the NRR plants noted if the generation from those plants exceeds net exports; if net exports are less than the production from those plants, the production from the plants is considered exported.  This is entirely theoretical - but the theory is that when these units are running they are the marginal supply and therefore the price setter - with exports reflecting customers who take that price.
[3] The figures used in the spreadsheet are $500/MWh for solar, $135/MWh for wind, an NRR of $13457/MWmonth, natural gas supply at $7/MMBtu, a heat rate of 7.5, and greenhouse gas emissions from natural gas-fired generation of 398 kg/MWh