The Uncle Gnarley Blog has a new website. Click here to visit www.unclegnarley.ca to view the latest posts!

Monday 20 November 2017

LOW RESERVOIRS POSE GREATEST OUTAGE RISK SINCE DARKNL

Guest Post Written By PlanetNL


PlanetNL5: Low Reservoirs Will Strain Thermal Generation This Winter
Uncle Gnarley suggested readers be given a respite from the gloomy Muskrat analysis series and instead asked PlanetNL to provide a critique on a different energy concern: the impact of low reservoir levels on this winter’s energy supply needs, as just revealed in a recent NL Hydro report.

The first concern is the cost of unplanned high fuel consumption at Holyrood and the standby combustion turbines over the winter and possibly well into 2018, a cost that NL Hydro is certain to want to recover through rate adjustments.  A considerable pre-Muskrat rate hike will result. 

NL President Jim Haynes
The next concern is a more chilling worry.  Can the Holyrood plant and the turbines, mostly aged assets in need of extensive repair, withstand the biggest demands placed on them in many years?  If not, the entire island system is at considerable risk of major outages.

Last week, NL Hydro released it’s latest twice-annual “Near Term Generation Adequacy Report” to the Public Utilities Board.  As expected from a Nalcor subsidiary, they have given themselves a passing grade and conclude they have adequate capacity margin for this winter and for the next 4 years before Muskrat generation is fully on-line and even if mainland interconnection is incomplete that whole time.

Within the report it is revealed that in the July through October period this year, barely half the average reservoir water in-flow rate has occurred, resulting in low energy reserves at key hydro generation sites. Copied below is a Table from the report that presents those in-flows in their equivalent energy value.  The historic average annual value totals 4600 GWh but 2017 is coming up well short.
 The report also included this useful Figure copied below presenting that indicates reservoir energy level is approximately 1000 GWh lower than ideal and over 500 GWh lower than the 20-year average.
The Figure also shows the story for winter 2016-17 which was a normal season.  Above average late fall 2016 rains brought water levels to near maximum just before the high demand winter heating season, however, winter in-flows were below average, resulting in overall average inflow.  The reservoirs were drawn down very evenly from 1-Dec (orange line at right) to just after 1-Apr (blue line at left) by over 1200 GWh and still had some margin to spare when spring thaw started new in-flows.  As a result, Holyrood supply last winter stayed on plan.  The system was adequate, reserves were present and large power outage events were few and brief.
Recent photo intake at Hinds Lake (submitted)
Barring any major rain events this month, December 2017 will begin with over 1000 GWh less reservoir energy available than a year ago. The reservoirs are barely 300 GWh above the minimum storage target level.  The NL Hydro report acknowledges that Holyrood is already running harder than normal this fall to limit the drawdown of reservoir energy.  Glaringly missing from the report is a more detailed analysis of energy needs through the winter period.  Let’s do one here.

Winter Energy Requirements – December through March
The 2018 Island Interconnected System Energy Forecast included in NL Hydro’s General Rate Application filed this year, shows the energy supply required during the 4-month winter period totalling over 3200 GWh.  Holyrood is expected to produce 1300 GWh while other non-hydro sources add a little over 100 GWh. 
The balance of 1800 GWh supply is set to come from hydro generation.  As the historic average monthly inflows (see the Table above) total to 1300 GWh, therefore the drawdown from reservoir storage would be 500 GWh. But there are two circumstances that makes the problem appreciably worse.  One is that hydro turbine efficiency drops significantly when water levels are low (lower head pressure reduces power output).  The second is that NL Hydro acknowledges their weather forecaster, AMEC, has advised to expect below normal precipitation for the next few months.

The reservoir energy deficit is likely 800 GWh.  Ideally, an alternate source of supply would be switched on during the winter period to not push the reservoirs below their specified minimum target levels and avoid a carryover problem to next winter.  That would require 270 MW of steady supply for 4 months and it simply doesn’t exist.  Even if the Upper Churchill recall was available this winter (it is scheduled to operate by next winter) it would only deliver 110 MW of firm power and with the Labrador West mining industry possibly warming up again, it could be even less.
Thermal Generation To The Rescue?
NL Hydro must find a balancing act to use hydro generation judiciously this winter while maximizing use of it’s thermal assets – a group of plants that don’t inspire confidence.
The Holyrood plant is the only one intended to supply regular baseload power and it is already planned to run to about 80% of it’s maximum total theoretical utilization and couldn’t be asked to do anything more until at least April when loads appreciably drop off.
The energy deficit has nowhere to go but to the combustion turbine fleet that are intended to serve as occasional brief emergency backup rather than as baseload power supply.  
The 50MW turbines at Hardwoods (St. John’s) and Stephenville are practically obsolete, have a spotty service record, and planned major capital upgrading remains incomplete.  The 120MW turbine installed at the Holyrood site a few years back, while newer and more serviceable, is due for major inspection and overhaul next year.  If all three units are pushed into hard service and perform well, they will deliver energy of about 100 GWh per month.  
Assuming 400 GWh of energy is successfully generated by the combustion turbines from December through March, the remaining 400 GWh reservoir energy deficit can be produced through no action except by running the reservoirs well below their minimum target levels – presumably unprecedented territory.  Between April and November of next year, barring above average precipitation and in-flows, Holyrood is likely be called upon to produce the remaining energy deficit of 400 GWh, no trivial matter itself and not something the utility can wait out and trust to the unproven Muskrat transmission line to resolve.
Cost Consequences
Hydro generation is essentially a fixed cost business with no production related variable costs – pay the overheads and the energy is practically free.  Thermal generation has the substantial variable cost of fuel.
Holyrood burns Number 6 type fuel, presently forecast in the GRA at $86.68 (CDN$) per barrel average cost in 2018.  The plant’s energy conversion rate is 608 KWh per barrel leading to a fuel-only cost of 14 c/kwh.  The standby combustion turbine plants are both less efficient than Holyrood and they burn higher-price diesel: the fuel energy cost is typically double that of Holyrood, therefore 2018 costs might be 28 c/kwh. 
The 400 GWh of turbine generation computes to a fuel cost of $112M.  The 400 GWh assigned to Holyrood will add $56M in fuel cost.
NL Hydro will want to recover these costs from ratepayers.  Based on 7000 GWh total island energy sales, recovery of this unplanned $168M fuel cost would require a rate adjustment of 2.4 c/KWh to be passed on to all customers for a period of 1 year.  
If these costs are treated under the 5-year Rate Stabilization Program to defer and average exceptional cost items, substantial interest costs are added and the rate premium would be 0.5 c/kWh minimum over 5 years.  However, a 5-year period overlaps the enormous future Muskrat price hike increasing rate shock effects.   
Precedent and the PUB
The PUB will be confronted with several options should these costs come to bear.

The Board may deny NL Hydro part or any of the cost recovery if they find more prudent management could have reasonably avoided extra fuel costs.  As precedent, the PUB rejected a February 2016 request by NL Hydro to recover turbine fuel costs that winter which NL Hydro attributed to a similar low water problem.  As Uncle Gnarley summarized later that month, in a blog post titled “Incompetence In Real Time”, the poor state of maintenance at Holyrood was the root cause of the issue and therefore ratepayers should be spared the cost.
With the power of hindsight, it can also be surmised that the urgent low reservoir situation faced by NL Hydro in winter 2015-16 wasn’t truly urgent at all.  Reservoir energy was over 500 GWh better than today and was easily at or above historic average levels before December yet the application stated the energy deficit was an astounding 1200 GWh.  This level of estimating error seems to the Nalcor way and is not limited to the Muskrat project.  As further indication of possible subterfuge, while the historic water level average line is included in last week’s report, it was absent from the February 2016 application.

Unlike two winters ago, the problem of winter 2017-18 is substantial and real, and it is likely the PUB will have to fairly consider this situation.  Recovery of any approved amount is likely to be considered in tandem with other rate increases including the 0.9 c/kwh increase made July 1, 2017 and the proposed increases of 0.7 c/kwh July 1, 2018 and 0.6 c/kwh January 1, 2019 in the GRA presently under review.  The PUB may ultimately decide to do a highly customized 2 or 3 year recovery period to clear the slate before Muskrat rate hikes of an additional 5-10 c/KWh arrive by 2021.
Reliability Worries
The “Near Term Generation Adequacy Report” goes into some detail assessing available generation capacity and concludes that NL Hydro has adequate reserve margin and will be reliable under their normal assumptions for generation equipment availability.  At no point, however, does the report indicate that the three combustion turbines will be used for steady baseload power generation which will often reduce or eliminate the reserve margin needed to avoid major outages.

When those standby resources are pressed into full-time service, they are also at risk of suffering new and unanticipated failures that may lead to their derating or complete unavailability.  The same applies to the Holyrood plant whose ongoing life cycle extension program is a feat of maintenance engineering by the line management and staff tasked with keeping it running. 
NL Hydro needs to rerun its report with decreased hydro efficiency, lower hydro energy availability, combined with revised assumptions when using standby turbines for baseload power generation.  The highest-level stress cases in last week’s report already look bad but NL Hydro seems to dismiss them as unlikely.  It is quite probable that a properly revised model will show considerably increased risk of major outage events: the unlikely could be become the probable case.

While NL Hydro does have capacity assistance contracts, principally with Corner Brook Pulp and Paper and a few other far smaller sources, these provide for load curtailment or for Deer Lake Power to add capacity for short bursts of up to a few hours duration only and for only so many successive days.  These measures will help to a point but in the event of multiple long-duration equipment failures and a serious hydro reservoir issue, the benefits will be limited.
NL Hydro is scheduled to present their overall Winter Readiness Report to the PUB in early December.  Let’s hope that report will contain a much more thorough approach to energy reserves and operating philosophy and that the PUB will not hesitate to demand clarification if it comes up short.

Plan for Outages
The potential for breakdowns during high load periods will be the greatest in recent memory and certainly since the Dark NL event of January 2014.  Much of the present spike in capital program work at NL Hydro is being spent to address deficiencies identified in reviews (mainly by the Liberty Group) completed after Dark NL, however, plenty of the recommended work still lies ahead in the 2018-2022 Capital Plan.  This winter’s real-world stress test of a partly upgraded system is coming a few years too soon for comfort.
Is NL Hydro ready for major outage events? What is their operating plan to minimize outages? How and when do they intend to communicate the low reservoir problem to the public?  Is there a contingency plan for days when the reserve margin drops below desired levels other than saying listen for the “power watch warning” and asking the public to down their heaters?  Early notice to the public of a challenging winter ahead could result in their taking important individual actions to prepare themselves for difficult circumstances.

Likewise, is the Government ready to deal with the issues of vulnerable people at risk or will a different Premier soon say he does not see it is as a crisis about to happen?
For many of the public, this event could be their personal tipping point to act on getting their energy consumption down.  It is inevitable that this will happen before the full measure of Muskrat-driven rate increases occur in 2021.  Together with recent and projected rate increases, the possibility of a one-time fuel cost recovery charge could increase rates by about 40% in the space of 18 months since July 1, 2017.  Those who are contemplating switching off electric resistance heating, should hardly need more motivation.

Muskrat Implication
The Pro-Muskrat crowd – particularly the New Muskrateers, Ball, Coady, and Marshall – will likely attempt to spin the issue as a justification for Muskrat Falls and will laud the potential for mainland energy interconnection to prevent this from ever happening again.  Such boasts would be to ignore terrible past decisions, the possibly extraordinary price of imported power, and the remaining opportunities to reduce future hardship. Had the right decisions been made, an isolated island system could have withstood this low water event at a fraction of the fuel cost and with a miniscule capital program relative to Muskrat.  It may still prove to be the most sensible solution available if the elusive Muskrat can be killed.

References:
NL Hydro, Near Term Generation Adequacy Report, November 15, 2017: document not yet uploaded to PUB website or not located; obtained through public email release NL Hydro, General Rate Application 2017, see p.144 of pdf for 2018 Energy Forecast:


Uncle Gnarley post, “Incompetence in Real Time”, February 22, 2016: 

NL Hydro application to PUB, February 5, 2016, for turbine fuel costs arising from alleged low reservoir level concerns: