Thursday, 18 September 2014


Written By "JM"

The New York Times recently published an article "Large Dams are Just Not Worth the Cost".  The opening line foreshadowed the main message of the article:

THAYER SCUDDER, the world’s leading authority on the impact of dams on poor people, has changed his mind about dams
Now in his 80’s, and after a lifetime of promoting large dams, Mr. Scudder has had a Gus Etchegary type of enlightenment.  He is now preaching caution against an industry that he spent his life promoting.  He was compelled to go public with his revised assessment after reading the Oxford University study released in March of this year.

The Oxford Paper, written by economists Ansar, Flyvjberg, Budzier and Lunn, is well worth the $10 purchase price for any person who is interested in public policy based upon mega-hydroelectric development.  It is especially relevant for those who want to assess the record of the Muskrat Falls project in a global context.  

The paper reviews the record of large dam construction on the basis of cost overruns, schedule variance, and overall benefit to the economy of the host countries. It is rich in historical references, and global context.  Had the research been available during the Muskrat Falls debate it would have provided an evidence based comparison with the performance of similar projects and provided a much needed challenge of Nalcor ’s billing of the project.   

The Authors of this Paper have reached the following conclusions:

1)      3 out of 4 dams experienced a cost overrun.
2)      Actual costs are, on average, 96% higher than estimated costs; the median is 27% higher than the initial budget.
3)      The evidence is overwhelming that costs are systematically biased towards underestimation at the original budget.
4)      Eight out of every 10 large dams suffer a schedule overrun. 
5)      Actual implementation schedule is on average 44% (or 2.3years) higher than the estimate with a median of 27% (or 1.7years)
6)      Like cost overruns, the evidence is overwhelming that implementation schedules are systematically biased towards underestimation in the initial planning stages.

How does the current performance on the Muskrat Falls project compare to the findings from the Oxford Paper?  Considering that Nalcor made the decision to proceed with Muskrat Falls at DG2 in November 2010, the cost data from this period of time should be used as the baseline data.  The following table shows the evolution of the cost estimates.  This is the data Nalcor presented to the public with the recommended escalation and contingencies included.  Nalcor deliberately omitted inclusion of “Interest During Construction”.

Based upon the DG2 budget, there has been a 42% increase in that estimate. The figure is comparable to the conclusions reached within the Oxford Paper.  Of particular concern, the element of the project furthest advanced (i.e. the MF to CF transmission line) has more than doubled the 2010 estimate.  Is this a foreshadowing of what is yet to come?

Nalcor has acknowledged a 13 month slip in the schedule for production of first power since the project was first presented to the PUB.  The delay echoes the candid view of the Independent Engineer's November 2013 report which states: 

Nalcor and the Government of Newfoundland have tried to put a positive spin on the under delivery to date.  They have stated that the cost and schedule overruns are similar to those on other mega projects in Newfoundland and are the result of a white hot economy.  Yet, placement of some 700,000 m3 of concrete has only just begun.  There remains much further opportunity for cost over-runs and schedule delay, too.  We should remain concerned! 

The question ought to be asked: why were the over-runs not included in the original cost estimates in the form of higher contingency allocations?  The other resource projects, currently underway in Newfoundland, were certainly known in 2012.   Pressures on costs and availability were easily predicted by anyone familiar with the Newfoundland construction industry.  This risk was clearly identified by many in public submissions to the PUB.  So, why was Nalcor’s contingency allocation so low?   

The Oxford Paper has fascinating commentary that hydro dams often advance due to biased decision making by the proponents:

Our approach to address the debates about whether or not to build dams is to incorporate an evidence-based perspective that reflects how decisions among alternative options are actually made and on what basis.  Theoretical and empirical literature on decision-making under uncertainty proposes two explanationspsychological delusion and political deceptionthat suggest decision - makers' forecasts, and hence ex ante judgments, are often adversely biased

The Authors argue that experts (e.g., statisticians, engineers, or economists) and laypersons are systematically and predictably too optimistic about the time, costs, and benefits of a decision.  With the power of hindsight, does anyone doubt that the management of Nalcor fell into this trap in planning Muskrat Falls? 

The Paper argued that optimistic judgments are often exacerbated by deception, i.e. strategic misrepresentation by project promoters (Wachs, 1989; Pickrell,1992; Flyvbjergetal.,2002, 2005, 2009). Recent literature on infrastructure delivery finds strong evidence that misplaced political incentives and agency problems lead to flawed decision-making.

This paragraph succinctly summarizes why I was so initially opposed to the Muskrat Falls project!

First, Nalcor’s initial decision making was not driven by an unbiased and objective review of all options to meet our energy requirements.  In July 2010, Nalcor made the decision to change the scope of the Lower Churchill Project team from that of energy export (Gull Island) to meeting the domestic demand of Newfoundland and Labrador (Muskrat Falls first).  With this change, Nalcor did not modify its Gated Management Process to reflect the new mandate of meeting domestic supply.   

In reality the project should not have passed DG2 until all options were examined, and Muskrat Falls was selected based on the lowest cost criteria established by the rule of law (Electrical Power Control Act-1994).  This lack of compliance with the rule of law was presumably endorsed by the Government of the day.  The endorsement was either the result of indifference or ignorance of the legislative requirement for the lowest cost criteria.  

The Canadian Environmental Assessment Agency (CEAA), as part of the Joint Review Panel, was first in publically identifying the flaw in Nalcor's decision making process:

Following this damming recommendation from the CEAA, and public pressure by David Vardy, Ron Penney and others, there were efforts by Nalcor and the government of Newfoundland to validate the project to public scrutiny.  The following actions were completed:

1)      Independent Review by Navigant in September of 2011
2)      Reference to the Public Utilities Review
3)      Review by Manitoba Hydro of the Muskrat Falls project to meet decision gate 3, and prior to the sanction of the project.  This process did include a review of several alternatives to Muskrat Falls to meet our needs.

A second major objection to this project was the influence brought by either Nalcor or the Provincial Government to ‘ring fence’ the terms of reference of these independent reviews.  Whether it was a strategic misrepresentation or just an honest simplification, the various third parties were not permitted to examine the project as it was sold the public.  As outlined within Part I of this series, Nalcor has recently admitted that the reliability calculations, and lowest cost analysis did not include the delivery of energy to Nova Scotia.  The lowest cost option as verified by Navigant, the PUB, or MHI is not what has been presented to the ratepayer.

Nalcor should be held to task to explain why the obligation to Emera of the 167 MW Nova Scotia Block  was excluded from the various analysis.

Finally, as noted by the Oxford Paper, the proponents of the Muskrat Falls project were too optimistic on the benefits of the Muskrat Falls decision.  Whether it was the availability of power for Labrador Mining, the benefits of export revenues, or the closure of Holyrood, both Nalcor and the Government of Newfoundland and Labrador consistently overstated the benefits of the project.  Of course, they have also understated immense risks such as cost over-runs and the Water Management Agreement, as discussed within Part II of this series.

As Muskrat Falls enters its mid game, we should remain concerned that project costs will continue to escalate and that the schedule continue to slip.  The next estimate, anticipated in the Fall of 2014, will be an indicator.

Only time will tell if Ed Martin will one day reach enlightenment like Mr. Scudder.  I wonder if Mr. Martin considers a 42% cost overrun a success?   

Editor's Note:  
DELUSION AND DECEPTION (Part III) was written by "JM".  

He is the anonymous researcher and writer who presented a major Paper to the PUB Review during the Muskrat Falls Review.  JM has written a number of Pieces for Uncle Gnarley Blog, including, most recently, VISION BUILT ON DELUSION (Part II) and The Snow Job (Part I).  Others include: Gnarley's Theory of Political Devolution and The Great Revolutionary From The Shore and The Right Side of History

JM has also submitted to the PUB a Paper entitled:  Underestimating Peak Load and The Potential Impact On The Muskrat Falls Solution

- Des Sullivan


  1. In your earlier piece you wondered if Nalcor's misrepresentation was strategic or deceptive.
    Consider David Vardy's question no 4 to Nalcor : Can 900 Megawatts be guaranteed over the Labrador Island Link? Nalcor's answer: The link (HVDC System) will be capable of transmitting 900 MW from Muskrat Falls at any time during bipole operation. During an outage on one pole the link will be capable of transmitting 900MW for a short period of time and tehn ramped back to 675 MW. Transmission losses are fully accounted for in all of Nalcor's calculations.
    One might believe that the answer to Vardy's question is YES. But that is not so.
    Vardy's question was whether 900MW could be transmitted over the link. This link transmits from Muskrat Falls to Soldiers Pond on the Avalon. A reasonable person would assume from the question and answer that , yes, 900MW arrives on the Avalon. That Vardy's term "over" implied "gets to" the Avalon. But that is not so.
    The truth is that 900 MW leaves Muskrat Falls and transmission losses absorbs a significant portion along the lines. When transmitting 900MW, 92.1 MW is lost allowing 807.9 MW to arrive at the Avalon. These are the figures Nalcor has provided the PUB.
    So is not Nalcor's reply to Vardy deceptive? WA, Logy Bay

  2. Low transmission losses is one of the advantages of having generation sources close to the load, the user of electricity.
    With 92.1 MW transmission loss on the Link, in normal operation, that is 10.2 percent loss
    With one DC pole out, the transmission losses jumps to 272.8 MW, as the single conductor is very overloaded. This will continue for 10 minutes until they ramp back the power being transmitted, reduced to 675 MW leaving Muskrat Falls and 530.6 MW arriving Soldiers Pond, for 144.4 MW transmission loss.
    Vale Inco at Long Hr will use about 80 MW. For comparison, Under normal operation, in bipole the losses on the Link, at 92.1 MW exceed Vale Inco usage. Under an outage of one pole of the link, transmission losses peak to almost 4 times that of Vale Inco usage.......enough to generate our own Northern Lights!!!!
    Transmission losses from Holyrood to St Johns is about 2 percent, say 10MW when delivering 495 MW.
    Conservation, especially efficient residential heating, heatpumps, reduces the load for the customer, about 3 kw per average house, 300Mw reduction for 100,000 houses. This reduces transmission losses, about 30 Mw of saving on the grid, real energy saving that can be sold to other customers. Of course this is Demand Side Management that is being ignored. WA

  3. Thank you JM for your work but you are letting Mr. Martin off too easy. What about interest during construction. It is a real cost. Workers and contractors on site this week must be paid with borrowed money and there is no revenue for several years so it must be added to the capital cost. The interest expense during construction is no different than money paid to any other supplier. It must be paid or there are consequences.
    If the interest wasn't part of the DG2 estimate then Nalcor low balled that estimate. Adding the interest cost brings the overrun into the 60% range. Also I believe the DG2 estimate had a contingency. Is there a similar one in place now.
    The province needs to start planning for possible poorer outcomes from this thing such as losing the water management rights court case, further overruns combined with a few years of lower oil prices driving up deficits and the cost of capital for both operating and MF.

  4. There had been a long term trend of more and more residents using electric heat as the primary heat source. It was assumed this was would continue. I am amazed to learn than since 2010 there has been a big reduction in the percentage using electric heat as the primary source. Oil use has increased and a large increase in wood, especially off the Avalon. And the high cost increases are yet to come, so people are switching in anticipation of high prices or unable to afford electricity at current prices. This trend does not bode well for the MF vision.