Well look at that. It's the end of December, and you know what that means...
It's TWO YEARS since the Manitoba Hydro Whistleblower made a complaint to the provincial Ombudsman.
Two years and still no resolution.
Resolution? Hell, two years and no nothing.
Resolution? Hell, two years and no nothing.
Ombudsman Irene Hamilton has managed to botch the first complaint under the NDP's much vaunted Public Interest Disclosure Act so badly it will never be used again.
Now there's a legacy.
First Hamilton tried to pawn the complaint off to the Auditor General, who, oops, turned out to be a former member of the Manitoba Hydro board of directors, and who announced defiantly she was under no obligation in law to investigate anything.
When she eventually acknowledged that the perception of conflict of interest would render moot anything she did with Hydro, the complaint was returned to the Ombudsman's Office.
Hamilton then decided to wait to see the results of a report commissioned by the Public Utilities Board into Hydro's risk management, which was to include an examination of the whistleblower's concerns. That 350-page report was released in November.
Since the Ombudsman is now entering her third year of inaction on the file, we decided to see for ourselves how the Hydro Whistleblower, who we've affectionately dubbed Miss Whistle, fared in her latest review.
Yes, folks, we read the 350 page report over the Christmas break. (Yeah, yeah, we know---get a life.)
* Miss Whistle was hired by Manitoba Hydro as a risk consultant following the disastrous drought of 2004.
* She concluded that the utility lost millions of dollars unnecessarily, and ratepayers got dinged with unwarranted increases in their bills, because Hydro mismanaged the province's water resources.
* Needless to say, this observation didn't sit well with the bosses who were responsible for running the utility. So she was shown the door.
* That's when she took her concerns to the Ombudsman under the Whistleblower Act.
* We're finally able to pinpoint many of the problems she saw within Manitoba Hydro operations. (All quotes, unless otherwise stated, are from Manitoba Hydro Risks: An Independent Review, by Dr. Atif Kubursi and Dr. Lonnie Magee, submitted to the Public Utilities Board of Manitoba, Nov. 15, 2010)
1) Topping the list are the "suite of models" that Hydro uses.
"Manitoba Hydro supports, uses and relies on three major models (HERMES, SPLASH and PRISM) in its planning of operations, investment planning, financial forecasting and budgeting."
Miss Whistle said the in-house models use outdated programming and stale numbers which lead to poor decisions on water levels, pricing, and the risks Hydro is taking.
She was specific in a 2009 letter to the editor of the Winnipeg Free Press (which they refused to run but which you can read in The Black Rod):
"To keep the lights on in the Province, Hydro’s computers rely on a complex formulae, which operates not just the reservoirs like Lake Winnipeg and Cedar Lake, but of course looks at the import tie-lines from the US, and Hydro’s ability to the run expensive gas units."
"While it may be shocking to learn these problems exist, what was uncovered was there were “systemic and massive” computer system flaws - with obsolete computers maintaining the calculations. Massive system errors and inadequate mathematics were found in the power calculation of “blackouts” or reliability conditions - which could lead to faulty results in keeping the lights on."
"This could be seen as the equivalent of using rotary phones in the age of wifi and Bluetooth technology."
2) The McMaster profs said:
"It seems that both Quebec and BC have adopted more advanced systems than MH’s Hydro Electric Reservoir Management Evaluation System (HERMES) or Simulation Program for Long Term Analysis of System Hydraulics (SPLASH)."
And...
"...MH models are serving their purposes and can be relied upon for operational planning and long term planning, but they need to be upgraded to include stochastic and dynamic modules. They also need to be reviewed and authenticated by external subject matter experts. It is desirable and beneficial to initiate formal documentation and integration of these models on a common platform."
Translation: the models are okay for what they do, but everything about them should be changed to keep up with the times, starting with what Miss Whistle said.
3) She was horrified to learn that Hydro's models were operated like a "black box"--- they used it but nobody could explain how it works.
" If it was learned, that the safety methods for keeping the lights on were programmed only by 1 or 2 persons (with source code and changes known only to them), on an outdated computer which hasn’t been changed since the 80’s, I think you too would be worried." she wrote.
Miss Whistle said the ICF consultants hired by Hydro to discredit her were told not to look at the computers. But the McMaster profs did.
"MOST is a decision support system (DSS) for hourly scheduling of generation and reservoir operations for the entire Manitoba Hydro generating facilities and transmission network.
Findings
A number of questions about MOST (Vista) and a number of recommendations emerge:
A number of questions about MOST (Vista) and a number of recommendations emerge:
* More than one skilled person should be responsible for the model."
And....
"Model Output: Lake Balances
Findings
Notwithstanding the small dollar amount of discrepancy between the Generation Estimate and HERMES solutions, these discrepancies raise concern about the accuracy of the model and the reporting system.
But the real problem is more profound.
HERMES and SPLASH are static models and do not handle time in a manner consistent with dynamic programming. MH may wish to consider some of the existing dynamic programming systems in use at similar utilities in North America."
*** The profs also picked up on Miss Whistle's red flag over Hydro's risk management practices.
"The qualitative aspects of risk management are well in place at MH. Unfortunately, this is not the case when it comes to the quantitative areas of risk management. There is hardly a mention of the word “quantitative” in the CRMR. Risk management is ultimately about quantification of exposure and calculation of the magnitudes of losses and threats."
4) Miss Whistle suggested Hydro "incorporate a mark to market framework for monitoring risk capital exposure and profit and loss optimization" (as Hydro put it in a secret internal document).
Hydro scoffed at the idea and said they use good old Canadian accounting principles, not fancy American ones.
The McMaster report concluded:
"The quantification of financial exposure should use fair market values (replacement costs). The Mark to Market (MTM) measures shall take precedence over other benchmark evaluations of financial risks."
Score another one for the Whistleblower.
5) But where Miss Whistle lost points with the McMaster profs was over her concern with Hydro's plans to lock in long term sales contracts with American buyers.
She recommended that Manitoba Hydro concentrate on building up a big war chest to tide them over any future drought without worrying about blackouts. She said Hydro could make more money from selling power on the spot market. And the danger of a drought long enough to eat up Hydro's reserves, before forcing the utility to default on foreign sales and its outstanding debt, wouldn't pass until about 2018 when more power plants were built and Bipole III was in place to carry load south.
"...there are no grounds to believe that there exists a serious material risk for blackouts in Manitoba." they wrote.
Note the weasel words:
Grounds to believe.
Serious.
Material risk.
That's not exactly saying there's no risk for blackouts, is it?
6) And seeing that the Whistleblower is basing her concern on proprietary software, which is more advanced than Hydro's "suite of models" you might expect the economist duo at least to reference the defects in her analysis. They don't, because they never got to examine how she reached her conclusions.
They instead depended on Manitoba Hydro's sweet spiel. Translation: keep your fingers crossed.
Hydro, said McMaster, needs to sign the Yanks up now or lose the sales to wind farms that are going to be built. (Honest. We're not making this up.)
And the deals require the American customers to build a transmission line to the border, giving us access to the U.S. market at no cost. We get it wholesale, see.
But one big reason to bless the long term contracts stood out for the profs.
"Finding 14
But the major achievements in these contracts are the curtailment provisions in the new contracts that could effectively decrease MH’s firm export commitments by 29%... in times of adverse water conditions."
Translation: In the event of a really bad drought, Manitoba can send as much as 29 percent less power to the United States and still be in their good graces.
7) Our dedicated team of codebreakers brings you the following Manitoba Hydro secrets:
-- In its deal with Northern States Power (NSP) of Minneapolis, Manitoba Hydro can reduce guaranteed energy during winter (Nov. 1 to Apr. 30) in 50 Megawatt increments, provided we give them notice by mid September. Manitoba Hydro still has to pay a penalty of the difference between the contract price and a complicated-formula price for natural gas.
-- In its deal with Wisconsin Public Service (WPS), Hydro can stop sending power on weekends if it can't meet firm energy commitments.
-- And in its deal with Minnesota Power (MP), Manitoba Hydro can, same as with the other buyers, reduce the electricity it sends them, but with the provisio it "must exercise at least 8 years of median or better water flows during the contract period" or else the contract is extended until that 8 years is experienced.
While approving Manitoba Hydro's series of mega-projects, the McMaster profs still slipped in some cautions of their own which carry echoes of Miss Whistle.
"... Long Term Contracts are now well structured and include many new and innovative curtailment provisions, reasonable escalators and upset heat rate based import prices. But these contracts need to be staggered over time and diversified over a larger group of counterparties."
"... the real material risks that can have drastic impacts on MH are financial. ..We argue strongly in favour of avoiding targeting massive borrowing, the debt structure of MH is already high and moral hazard behaviour must be avoided."
Seeing as how the PUB report was prepared by two McMaster University economics professors, we thought it appropriate to give Miss Whistle a grade based on their valuation of the validity of her complaints.
The grade: B+.
The grade: B+.
"The Consultant’s contributions to the improvement of risk governance, oversight and back testing of MH models, the quantification of risks, the use of stress tests and the need for provisioning of adequate risk capital are undeniably valuable and have prompted MH to rise to the challenge and to introduce several improvements in its operations and models," wrote the profs.
Ombudsman Irene Hamilton gets an F.