Market Commentary

Ansergy has added a new report, Daily Control Area Forecasts, that computes the following control area  metrics for daily on peak and off peak:

1. Load – Ansergy forecasts its hub-level load forecasts at a control area level
2. Hydro – Hydro is forecast at a Basin level and ownership of each basin is based upon the EIA 860 reports.  Basin level energy is split into control area using the relative ownership levels
3. Wind – most wind is under contract and only the wind that is reported as owned in the EIA 860, or is publicly controlled under long-term contract, is allocated to the control area.
4. Net Load = Load – Hydro – Wind
5. Change in Net Load = Today’s Net Load – DAYn
   1. This report acts as a proxy for a control area’s net exposure to the market.  The basis for this is that if we know their net position today their future net position will be driven by changes in its load, hydro, or wind.  Of course this assumes that the control area did not add new generating capacity and/or new long-term contracts.

The following map summarizes the Ansergy Control Areas and the hubs they belong to:

The following shows Puget’s  change in net  load for the off peak hours:

The utility will find itself getting shorter over the next few months and, as winter wanes, that short position becomes length with the ramping up of the runoff and warming spring weather.  Drill down into the elements (Loads, Hydro, or Wind) of the forecast to understand what is driving this:

As you can see the loads account for most of the swing in net position; the utility realizes an 800 MW increase at  winter peak and a 1400 MW drop in the midst of the spring doldrums.  Netting all three together returns Net Load:

The Change in Net Load is computed by taking today’s net load and subtracting that value from each of the forecast days.   If the analysis is missing something today (wind, load, contracts, hydro) that same value is missing from all the future forecast dates and thereby rendering the Net Change in Load a reasonable proxy for how a utlity’s position is changing across time.

If you have any questions feel free to contact me.

Best Regards,

Mike

Good Morning,

The Mid-C 7 day forecast is looking wetter than yesterday’s, which was an already wet outlook:

This is production-weighted precipitation and is still skewed by massive west-side rains:

This kind of event will cause flooding on some (maybe all) of the west-side rivers.  Though the reservoir owners will try to catch as much as possible to mitigate those floods the overall production in these basins will increase.  Furthermore, this storm will jack flows on the east of the Cascade rivers (Methow, Wenatchee, etc) which enter the Columbia essentially unregulated.  This won’t be a big driver but it’s bearish now matter how you look at it.

I’d expect the NWRFC to bump their 10 day today and most definitely we will see Monday’s STP raise the bom and perhaps Dec.

Mike

Good Afternoon,

Check out the plot of projected Mid-C precipitation, it’s gonna get wet (and stay warm):

That’s a lake-load of water that will be fall across the Mid-C over the next week; since it is within a week the odds of it actually occurring are good.  On the surface 0.3″ of rain doesn’t seem like much but keep in mind that this is weighted by production plus you should be looking at cumulative over the period.  In that vein the total precip is closer to an inch, or about 5% of annual totals.    I think the RFC is seeing the same thing as their most recent forecasts have surpassed Monday’s STP for the balance of the week:

No matter how the hydro look is sliced, diced, and digested there is only one logical conclusion ….. bearish.  Bearish for the BOM, slightly bearish for the near term.  Speaking of bearish check out a new report  Ansergy will be adding next week:

This report summarizes 131 fundamental metrics for each hub in terms of bearish/bullish by comparing the current forecast or most recent historical with yesterday’s or last week’s.    All of this data comes directly from Ansergy’s Fundamental reports.  A sample of the detailed report that drives the above table can be seen below:

This is a truncated copy of today’s report, there are 131 hub-level metrics used.   This report will be available as either an email update or a standard Ansergy report next week.

More bearishness can be harvested from the renewables as the Mid-C and SP15 charts testify:

Throw on the ramping up Diablo unit, and the PV one is not far behind:

Tempering all of this fundamental softening is the load outlook.  The following three plots show Mid-C, NP15, and SP15 actuals:

Peak loads are ramping up in the northwest, and even at NP15, but are still falling at SP15 which has now nearly reached the bottom of its seasonal load swings.

So am I a bear?  Short-term yes, longer term I am bullish on Mid-C, bearish on the ISO, bearish on Q2, neutral on Q1, and bullish on Q3.  I’d still wager that the Mid-C’s water year ends up below normal, probably in the mid 80s, and California has a near record water year.  The currently El Nino index (3.4) published a 2.8 last week (versus 2.6 same week in 1997) to set an all time high.  The current weather patterns are wet northwest and relatively dry California but I am expecting that to change starting around mid-December, though there is plenty of time to make an El Nino trade(s) given the very bearish near-term outlook.

A couple of trades that I’ve been watching are the Q2-Q1 rolls as I see Q2 bearing the brunt of El Nino induced bearishness (wet California with its current renewable mix could find the ISO exporting to keep its renewables online).   The Q2-Q1 Mid-C off peak is developing a potential opportunity to go long the 1 and sell the 2:

I like the on peak equally well but might wait for Q1 to get further burned by Nov weakness:

Note the market has found safe harbor by trading the roll dead center between the high and low historical settles whereas Ansergy is putting the roll closer to the high end.   Timing wise I’d wait as the front is more likely to come off more until the Mid-C sees some real cold which is not evident in today’s forecast.

Mike

Good Afternoon,

The NWRFC has completed it’s weekly forecast of northwest hydro flows.  The following plot summarizes the last five RFC STP forecasts.  By comparing the current week’s forecast to the previous four forecasts we can draw inferences regarding the validity of the forecast and its possible impact on the market.

Pretty chart, but what does it tell us?  For starters, our prediction that a cut in BOM has come true, though the change is small.  We also were hammering the RFC about shaping more water in Nov than Dec; as we can see from above they were listening and are beginning the process of backing Nov down and ratcheting up Dec.

But its still not right as we’d be more satisfied seeing the block of hydro energy shaped more closely with loads.  One other consideration is draft, which begins in February.  The current forecast shows Nov 1300 aMW over Feb, something we find absurd and leads us to believe both Jan and Feb will be subject to further up-revisions and BOM to be walked down further.

Another way to test validity is compare the current water year to prior years:

This most recent forecast is now looking as it should – lower in 2015 than the previous three years – because 2015 reservoir levels are lower.  A final test is to examine the daily energy, which is my favorite view:

This illustrates our issue with the overall shape of energy – the downward sloping shape of energy – we believe it should be upward sloping, more like our demand forecast for Mid-C:

Griping aside, the forecast is what it is, until it isn’t.  I believe we will continue to see water shaved out of Nov and pushed into Dec-Feb with the caveat being we don’t know weather.  If it rains like hades over the next two weeks maybe that got it right, but doubt it.

Mike

 

Good morning,

We had a wet weekend in Seattle and that trend appears to be continuing on through the rest of the week with at least another inch expected by week’s end:

Check out yesterday’s (Sunday) forecast for the 12th – 2.24″; apparently the A Team was home watching football and the weather scrubs were turned loose and ran amuck.  The pros backed that forecast down to 1″ but spread it across two days.    This storm track is clearly on a northern bent as California is getting very little, even Portland is missing much of the rain:

Clearly less precip in week one but check out those crack-cocaine forecasts for Thanksgiving week … nearly 4″.    Highly unlikely, too far out in the forecast, not even sure why any one would dare to place such an anomaly in a forecast (oh that’s right, Ansergy does it nearly every day).

Most of the northern basins are starting to accumulate snow, though they remain below normal.  The Clark Fork is at 50%, Salmon at 87%, Kootenai is at 67%, and the Flathead is at 76%.  It is far too early to call the water year but the trend is setting up for dry.   A study done a few years back showed the correlation between the current anomaly and final anomaly not becoming statistically relevant until mid-December.

The northwest reservoirs begin the water year nearly 4% behind the last three years:

All of this shortfall is explained by Libby and Hungry Horse (each off about 10%) while Coulee,  Dworshak and Lake Pend Oreille are near normal.  None of those facts can lead to any compelling water year trading/hedging strategy yet, they are just worthy of our attention as the year plays out.

Today will deliver the latest installment of the NWRFC’s STP forecast.  Probably the best tool to forecast the forecast is Ansergy’s 10 day versus STP report, at least it is in my opinion:

The trend from last week was to cut water in the 10 day forecast (for the same days as the STP) which would lead me to induce a modest cut in today’s 120 day forecast.   This induction is confirmed by relatively dry east-side forecasts, though if the RFC places any credence in Portland’s 21 day precip forecast they may go the other way.  This morning’s update will be important to monitor as it probably will tip their hand as to the STP’s direction.

Renewables have tapered off considerably adding a modest bullish element to the picture:

Most of the cuts in wind are coming from the Mid-C.    While we are on the generation front it’s worthy of noting that the Diablo Canyon refuel is over and that unit has begun ramping up.

Ironically, as Diablo 1 was returning CGS burped, but appears to be a short-lived phenom and now is almost back to 100%.  We are forecasting PV 2 to exit its refuel by week’s end thus returning nearly 2000 MW of baseload to the bearish WECC outlook.

Loads?  Slightly bearish in the south and neutral to slightly bullish in the north.  The weather forecasts show a hint of cold weather east of the Cascades at Boise and Spokane, but it is short-lived and not that cold (1-2 days, 15 degrees):

I’d put as much faith in those low temperatures as I did with Portland getting 4″ of rain on Thanksgiving, but it’s worthy of watching.  The Mid-C continues to fill the AC into NP15, and Powerex continues to remain, mostly, on the sidelines:

Clearly the folks up north are opportunistically buying and selling as they see fit which suggests a further dampering of MId-C price volatility.  If things get bullish they have bullets to sell; if prices get soft they will buy.  Meanwhile the northwest will keep the lights on in California and fill the AC, at least until Diablo comes back, then let’s revisit that premise.

The Ansergy price forecast for Mid-C remains range driven as the water year swings from dry to wet, but that should change as we get a bit further into the developing water year:

So given all of the above what are my sentiments you may be asking?  Don’t really have any strong opinions, now is the time to watch and wait.  I’d not be making major bets on wet/dry, just looking for market mistakes.  Tomorrow I’ll identify a few of those mistakes, at least what I consider mistakes.

Watch for our STP update later today, there might be some fireworks in that water report, but probably not.

In passing  I would like to mention with deep sadness the passing of Ryan Hoppe, a close friend and colleague of mine.  Ryan was my first hire at PowerLytix and most recently was in charge of Analytics at Tacoma Power.  He will be dearly missed by many.

Mike

Good Afternoon,

Just a quick update on a few things, more NW hydro focused than anything else. The last storms have now set a few of the principal basins in positive territory for precip, including the Clark Fork, Clearwater; and several others are now at normal (Kootenai, Flathead, and Pend Oreille).   Maybe more interesting is the recent rains on the east side of the Cascades have pushed the Methow and Wenatchee to 160% of normal.  These latter two can are relevant as they enter the Columbia below Coulee and are effectively unregulated.

Though these side flows are not contributing much to total flows today, they can account for 100kcfs in flood stage, something that can, and has, happened in early Winter.  I am seeing another storm blowing through the northwest as evidenced in the Seattle 21 day forecast:

Today’s forecast is putting nearly an inch of rain on Seattle, some of that will make it over the hill, and what looks like another storm next week.  None of this is coming as snow, it’s too warm, at least not in the lower elevation northwest sites.  Lake Tahoe looks like it is going to get some early snow:

Check out that off the wall forecast of 8 degrees on the 18th of November – a bold and ridiculously low prediction so far out, but that aside, there is .5 to 1″ of precip coming with temps certainly in the 20s.  Skiing by Thanksgiving?

Renewables remain robust, right?

Right, they do, right they are bearish, so is the nuke outlook.  Diablo and Palo each have a unit returning this month.  So what is bullish?  Nothing, its bearish everything and everywhere.   The River Forecast Center has given back what it tooketh last Monday:

And she is consistently being consistent all week long suggesting an unchanged STP on Monday, though I’d be inclined to predict that they may put some more water into BOM and Dec.  I know that is contrary to what I’ve been saying, but there are new storms a coming, storms that have passed through, and water supply outlooks that have been revised upwards, albeit modestly.

I’d suggest those forecasts (water supply) continue to be revised upwards through next week casting a bearish pall on Q2 and  Q3.  Speaking of which, let’s look at these (Mid_C OnPeak), beginning with Q2:

Our forecast has taken a drop in the last few weeks while the market remains mired and unchanged in the low 20s.  Yeah, I know it’s El Nino and, by the way, the latest weekly index is now 2.7 vs 2.6 same week in 1997, but we just aren’t seeing the classic El Nino dry.  I’d be inclined to off load any Q2 length today thinking I could pick it back up cheaper in few weeks.  The Q3 is more interesting:

The market has tumbled, the forecast has slipped a bit, but the delta is non-existent suggesting a possible long entry point.  So yes, I am saying how about long the Q3:Q2?   Yes, that is what I’m trying to say, “How about that Q3-Q2 Mid-C roll?”:

I am almost giddy over this trade, from the long side.  The market has pounded it down while the forecast has dropped much less.  From an inflection perspective there is way more upside in the 3 than the 2; probably some of the current price is driven off of last year’s end of June lotto ticket, one that has a very low probability of recurring.  Put me down as bullish on this roll.

 

 

Good Afternoon,

The NWRFC released this week’s STP late this afternoon and, as expected and previously foretold this morning, is slightly bearish for BOM, neutral for the other months:

One can’t, or shan’t, be too surprised by the forecast given their 10 day energy rally over the last couple of days.  To put this in perspective let’s look at that 10 day:

The back end of this short term forecast is slightly bearish but most of the new water is added to the first three days – indicative of rain-driven stream flow surges.   This point is further illustrated in the daily plot of STP energy:

The shape looks flaky, at least to me, given the plateau of energy in the first half of November, and suggests that days 6-15 are just a wild guess on their part.  I find it laughable that BPA would allow a surge of water in mid-Nov given a forecast of mild temps in advance of who knows what in December.

The year on year puts that point into perspective:

All four months of the current forecast are trailing the previous four years for the same months, which is how it should be … reservoirs are lower now than they were then.   The only issue I have with this plot is the Nov/Dec split; I’d be more comfortable seeing Dec over Nov; it has been that way in most years, and should especially be that way this year as BPA husbands its scarce water.

So chalk up this week’s forecast as bearish front end of BOM and neutral the rest.  I expect more volatility coming.

Cheers,

Mike

 

 

 

 

Good Morning,

Northwest loads are drifting up on cooler weather; southwest loads are now mired in the trough of winter doldrums, but that is not what is relevant.  Water, lots of water, reached the northwest over the weekend, check out current river conditions in Washington state:

Those blue dots are above normal, black is near record; the northwest has not seen blue and black dots for a while, they are now.  Perhaps this is why the NWRFC has substantially jacked up its 10 forecast:

This is a material change from last week’s STP; Nov 1 is up nearly 2,000 MWs.  The 10 day bodes bearish to very bearish for today’s STP, though I struggle to see how a few inches of west-side rain can affect production as much as the RFC projects, not that it matters, the reality is not reality, it is what the govt perceives reality to be.  If the STP puts a gig in BOM its bearish, pure and simple, though none of this precip has found its way into the regulated lower Columbia:

The Dalles is unchanged over the last two weeks, so where is the new water?  I’d say most of it is in Lake Roosevelt:

The spike in elevation coincides with the rain, though it’s a bit of a head fake as GCL outflows were cut over the weekend, but still its a substantial boost to BPA’s cold weather arsenal and we have no choice but to call it bearish BOM.  The precip is long over due as the following water supply plots (precip) for the Clark Fork, Flathead, Kootenai, and Salmon illustrate:

Don’t bother running the SWE …there isn’t much to see, everything is near 0.  All of these important basins are at five-year lows for Nov 2, though that is meaningless since one good storm will push them above normal.

Renewables are adding to the bearish BOM outlook with both Mid-C and SP cranking out the free energy:

Though Mid-C is by far the biggest beneficiary with nearly 2000 aMW more wind than last week.   Cali will get another 1000 MW when Diablo 2 returns, perhaps late next week, with the Palo unit right behind.  Overall it just doesn’t look bullish, and the weather forecasts are not helping.  The outlook for Mid-C is just neutral, SP is bearish.

The only bullish item of note is the relatively dry 10 day out look for Mid-C, nothing is bullish at SP.  Another slightly bullish element at Mid-C is the recent uptick in exports:

I guess with all that wind there was no choice but send it south, though I find it interesting that BC Hydro set a 6 month export high on Nov 1 with nearly 1000 aMW sold into the Mid-C, strange since it corresponds to very high wind generation.

Ansergy’s short-term price forecasts have been drifting down at both SP and Mid-C for BOM:

Much of this drop can be attributed to gas prices but the fundamentals remain bearish at SP and at best mixed at the Mid-C.  Today’s STP forecast should be interesting, I will go on record as prophesying a reversal of last weeks’ cuts, or at least a partial reversal, but we’ll see soon enough.  Post STP I’ll take a look at the forwards and share with you that which I love and that which I despise.

Cheers,

Mike

 

 

 

Ansergy will be making some changes to its models and reports over this weekend. These changes are intended to make the model more transparent and improve the forecast accuracy. Please continue to send us your thoughts for improvements.

New Forecast Items

Nukes – The “Served by Stack” item in the dispatch report will be disaggregated further by breaking our nuclear generation in a separate line.
Market Heat Rate vs Model Heat Rate – Several users have been confused about exactly how we compute heat rate. To provide additional clarity there are now 2 heat rates: Market Heat Rate and a second heat rate named “Model Heat Rate” defined as:

Market Heat Rate = Power Price / Hub Gas Price
Model Heat Rate = Power Price / (Hub Gas Price + Transportation adder),

where the current trans adder is $0.25 for all hubs. The model heat rate, by definition, will always be lower than the market heat rate.
TTC – Now TTC used in the forecast as will be presented along with forecasted transmission flows. This will provide more transparency when the forecast is affected by transmission constraints.

Forecast Sensitivities

Sensitivities will now be run as full model redispatches as opposed to the post process methodology used prior to now. That means each case will result in redispatch including altering transmission flows. The range of senarios will be 2500 MW higher and lower than the base case which represents approximately 1 to 2 standard deviations. The Sensitivty report will also include the following new items: Gas Burn, Coal Burn, CO2, NOX, and SOX emissions.

Snow-Based Hydro Energy Forecast for Current Water Year

In August Ansergy began forecast water year 2016 by taking an average of flows from the EL Nino years. That change resulted in about an 80% water year for 2016, a bold assumption given that the water year had not even started. In hindsight we would rather let the anomaly be proven by a dearth of snow and are consequently switching to a purely snow-driven forecast for the current water year. This change requires we re-run all of the history back to August 1 to insure the historical forecasts are in snyc with the current.

Going forward Ansergy will base its water supply volume forecasts off of the current anomaly but assume normal for the balance of the snow build year (until April 1). Previously we assumed a “normal” that used the average flows from El Nino years. In other words the current anomaly will be weighted against 100% normal for the balance of the water year versus around 80%. This change is bearish for the Ansergy term forecast but is more in line with what we know today, rather than what we think next June will look like. The two versions would have converged around Feb-Mar if this water year is dry; if it is not the current version would have missed widely from actuals.

 

Good Morning,

There is not a lot to say about demand other than loads are creeping up at the Mid-C and drifting down at SP:

This isn’t new, therefore isn’t news, as the trend at both hubs has been in place most of the month.  SP has a a gig or two to shed before it lies relatively dormant, load-wise, for the next 4-5 months while the Mid-C is set to add another 5-6 gigs by Christmas.  The renewable front is neutral with respect to the SP-MC spreads with both SP renewables and Mid-C weakening of late:

One indisputable bearish element, relative to that spread, is the current refuels of DC1 and PV2, and the fact they will return in the next few weeks adding two gigs of baseload.  If you haven’t gotten the love out of your SP or PV length by now you probably won’t for a while:

The first plot is the current status from the NRC over the last 90 days while the second is the Ansergy refuel forecasts for the next six months.  Please note I used the zoom in feature to highlight the near term outlook.  It is worthy to note that after the PV unit returns there is not another scheduled refuel until next spring.

On the transmission front not much is changing, BC purchased a decent block from Mid-C yesterday but essentially is not aggressively active in the market at the moment.  The ISO has shrugged off the DC maintenance and is now importing a gig or so less than it was a few weeks ago; clearly it doesn’t need the energy that otherwise might be flowing on the DC.

More worthy of some commentary than the above are the developing hydro dynamics in the northwest.  Let’s start with the River Forecast Center’s 10 day versus Monday’s STP:

This is bearish.  The RFC has added a gig beginning on Oct 31 to the next five days and a half gig to the remaining two days and thereby nearly offset the cuts they projected in the most recent STP versus the prior week.  Normally I’d discount this but the northwest is going to realize a significant precip event over the next few days and this is driving the RFC’s revisions, and probably will result in a bearish STP on Monday.

Some of the key locations  I look at are Kalispell, MT, Spokane, WA, Boise, ID, and Portland, OR:

The storm track is shifted north which is why California is not getting much and Boise’s precip is below Spokane’s (very un El Nino like).  Most interesting is the volumes in Kalispell and Spokane: nearly an inch cumulative in both locales.   Given that Spokane is lucky to get 18 inches in a year a storm of this size, at this early date in the water year, is material.  I’ve also included the min temps which suggests snow levels north of 5,000 feet, rain below.

Rain is the one element that will wash away whatever bullish sentiments we have with respect to Mid-C.  This is not enough but it begs the question “Has a new weather trend started?”  Worthy of watching, clearly the RFC has its eyes glued to the event and we expect more energy in next Monday’s STP forecast.

Cheers,

 

Mike