Nalcor is vastly overstating the usable output from Muskrat Falls that the grid will be able to accept. By overstating the benefits of Muskrat Falls, the utility is understating the rate impact.
Nalcor has spent approximately $500 million trying to develop a justification for the politically driven but fundamentally uneconomic project. For $500 million one can buy a lot of fancy legal, electricity, and financial tricks but all the smart help in the world can’t turn this dog into gold.
My argument is presented in a series of modules with some bolded sections to help you follow the thread.
Framing the Question
Here is a simple question: What is the incremental stand-alone per kWh cost of Muskrat Falls power delivered to Soldiers Pond near Saint John’s? Good question. I hope some day to meet the author who put this to Nalcor in the hearing. Everyone in the province should care about that question and everyone in the province deserves a simple, straight answer with all the calculations laid clearly out.
The first sentence of the reply is a preëmptive legal strike by the litigation team. To keep the real cost of Muskrat Falls power concealed, Nalcor has to ensure that this question is stopped and any supplemental questions that intervenors or the Board might raise also get stopped. The first sentence argues that the question is irrelevant. If pressed, the lawyer in the hearing will argue that even if the question was relevant (and they will say it is not) the utility has already provided a full three paragraphs of detailed explanations and a pretty picture. If pressed further, the lawyers want the option to argue that forcing an answer to the question would be unfair.
Although estimates for the delivered cost of Muskrat Falls are best considered on an integrated system basis, once an integrated scenario is established, it is a simple matter for Nalcor to calculate the cost for the Muskrat-plus-transmission element of the new integrated system. There is no justification for Nalcor not answering the question with full disclosure of the methodology and assumptions used.
Here is another simple question: How much power will Muskrat Fall generate?
Obviously, you can’t figure out the costs without knowing how much output benefit you’ll get.
Nalcor has been answering that question over and over. The reply always given is 4.9 TWh/yr of electricity.
Here is Nalcor explaining in June of last year that it conducted a long list of studies on exactly that question using Canada’s top engineering consultants. Nalcor puffs up the answer by pointing out that the consultants even used computers to perform the calculations.
The average monthly water flows on the Churchill River at Muskrat Falls are provided in the Environmental Impact Statement (EIS) here at page 5-5. The first observation to focus attention on before considering the units of measure is that the water flow, and hence the potential power production, is twice in June what it is in January. The usage pattern of power by consumers on the island is more or less the opposite — low in June, high in January.
The Muskrat Falls dam design has almost no storage, so inflow timing drives the power potential in nearly real-time.
I have manually calculated the area under the curve presented in the EIS to get a ball park estimate of Muskrat Falls production assuming no grid constraints and no down time for servicing.
The EIS says that the total max discharge rate from the powerhouse will be 2,660 m3/s. I have assigned this output as 824 MW.
My area under the curve estimate of the average production rate over the year is 577 MW (taking into account the nameplate capacity). Assuming a theoretically perfect 100% load factor, this corresponds to 5.05 TWh of production — i.e. pretty close to the project estimate of 4.9 TWh of production. Are you starting to smell the rat?
We already know that significant on-island generation potential is spilled due to differences between the patterns of natural inflows vs. the annual pattern of electricity usage. This is a normal and expected outcome for a system that gets a high fraction of its supply from hydro-electric generation. If the timing of inflows matched the timing of demand, the island wouldn’t need much or any oil-fired Holyrood output today. In an integrated system where Muskrat Falls is connected to the island, water coming to Muskrat Fall in June is useless for island consumers because the island generation is already spilling water. Generators closer to the consumer have reliability and cost advantages over remote generators.
During times when the island generation network without Muskrat Fall would be spilling, the 500 MW interconnector with Nova Scotia will be used to capacity, with that generation met from surplus on-island generation and the rest from Muskrat Falls. Any extra water at Muskrat Falls at these times will be spilled. In years of good inflows on the island and with the prospect of further loss of some significant industrial loads on the island, it is conceivable that Muskrat Falls generation will have periods in June when it will be generating nothing or close to nothing.
What about winter?
During the times of the year when Holyrood would have been running near capacity, all on-island generation will be running flat-out (as it does today) and Muskrat Falls will not be able to supply enough power to move any significant amount of power to Nova Scotia. This is because in Jan./Feb./Mar. the maximum output of Muskrat Falls will be about 500 MW due to the seasonality of water flow. This maximum output is after taking into account the operation of the Upper Churchill facility upstream which is contractually bound to maximize winter production for sales to Quebec.
No wonder Emera hasn’t signed the final deal. Emera has figured out that Nova Scotia gets nothing or close to nothing when customers there need it most.
Power exports from Newfoundland further downstream of Nova Scotia at peak winter times are absolutely zero.
What does this mean?
The key point here is that the annual output of Muskrat Falls is never going to get close to 4.9 TWh. Nalcor’s use of the term “firm” with reference to the output estimate of 4.5 TWh is also deliberately misleading. Muskrat Falls output will never get anywhere near Nalcor’s “firm” production estimate.
If pinned down, Nalcor’s regulatory economists might agree that some water at Muskrat Falls might get wasted but it will be insignificant. The power engineers might even have prepared for this threatening eventuality in the hearing by drafting some study that shows low water wastage at Muskrat Fall. If there is such a study up Nalcor’s sleeve, one thing it won’t include is an integrated system perspective on water losses. One can inflate the apparent production at Muskrat Falls by playing the mind game of taking all of the power spill that will happen on the integrated system and assigning it to existing on-island generation.
The actual usable output of Muskrat Fall will be a fraction of the 4.9 TWh Nalcor claims. I wouldn’t be the least bit surprised if the usable generation from Muskrat Falls calculated using realistic estimates for load, generation servicing, transmission servicing, and transmission congestion turns out to be half of the unconstrained potential generation that Nalcor is now parading around.
Although Nalcor knows how much the station will have to turn down due to grid constraints, the utility says that the answer to that question is confidential. (MHI-Nalcor 74 and 77)
Lower usable generation benefits from the investment increases the delivered cost of Muskrat Falls power.
When calculating costs, Nalcor (incorrectly) says that you may only look through the wide-angle lens showing the integrated system. However, when calculating production benefits, Nalcor’s says you may only look through its pin-hole perspective that isolates Muskrat Falls and ignores the constraints imposed by the integrated system.
You need lots of high priced consultants to cook up a regulatory mind game that sophisticated.
If the argument presented here is wrong, I will promptly issue an apology to Nalcor, its lawyers, and consultants.
A further analysis attempting to put the costs of Muskrat Fall in perspective will be presented on this web site in coming days.
Post script #1 (6:30 am Jan. 30): The above essay attracted the following important criticism: “Muskrat will be connected to Churchill Falls. Related to the CFLCo contract I think that CFLCo/Nalcor can wheel 300MW through Quebec but is currently doing a bit less than half of that. I think some of the Muskrat June supply could be swung through the Quebec grid (at least up to the 300 MW.)
The transmission connection between Muskrat Falls and Churchill Falls would be a factor improving the operational value of Muskrat Falls. Failing to address the benefits of this element of the project is an oversight in my essay. Quantifying the benefit of the MF x CF transmission link for the production output of Muskrat Falls would require load flow analysis for Labrador’s connection with Quebec and the transmission interfaces in Quebec flowing south.
While my essay should have addressed the benefits for Muskrat Falls of it being connected through Quebec, I do not believe that this factor has a large impact on the economics of Muskrat Falls. The reason is that Muskrat Falls has the same basic inflow characteristics as compared with all the other generators in the electrical neighbourhood — Churchill Falls, the Manic complex, Saint Marguerite, Bersimis, Peribonka, Toulnustouc, Outardes, and others. The soon to be completed 1550 MW Romaine complex will require significant capacity the the transmission carrying power south out of the region. Much of the hydro-electric capacity in Quebec sharing the transmission lines running south are run of river designs. Looked at as an integrated system, the generators in Labrador and north eastern Quebec will be spilling water during the freshet and leaving empty headroom on the transmission lines running south the rest of the time.
I wouldn’t be surprised if all the generators in the neighbourhood spill during the freshet.