My first article demonstrated clearly that the crash in the natural gas market (UNG, GASZ, GAZ, UNL, NAGS, BOIL, etc...) experienced in the spring, summer, and fall of 2012 was certain to not be sustained. In the discussions following that article, it became clear that there is some confusion about the relationship between wind power (FAN, PWND) and natural gas (NG) demand in the power sector. There seems to be a widespread belief that higher natural gas prices are good for wind power, and vice versa. Nothing could be further from the truth. The reality is that high wind penetration in some cases will result in higher natural gas prices. By spring of 2013, this is certain to be the case.
Natural Gas is crucial to have as the "balance power" (dispatchable backup) for wind power - which would otherwise be curtailed if there is too much reliance on baseload power (coal and nuclear). So the existence of low-cost natural gas ensures there is less integration problems, and therefore less risk for new wind farms or new expansions of wind farms. But the reverse is also true - as more wind power gets added to the grid, those same integration difficulties will change the price points for dispatch switching between baseload and natural gas (balance power).
As natural gas supplies become constrained this winter and storage levels start dropping lower than usual by spring; the price point for dispatch switching back to coal will be far higher than the price of coal alone would justify. Understanding this is critical to avoid losing a fortune in prospective gains as the wellhead price of natural gas climbs steadily towards and perhaps beyond $5/mcf by spring. So let's break this down further.
Wind power, growing fast enough to matter
The first criticism that this article will face is an immediate dismissal because "wind power is insignificant". That is an easy conclusion to reach looking just at national generation figures, with wind power generating only ~3% of the total electrical energy that was generated in the U.S. in 2011 (it will likely be ~4% of energy generated in 2012). But electricity is not a nationally distributed energy carrier. The electron impulses generated in Iowa can only be distributed to New York through great expense and high line losses. If you look at the states that have the highest penetration of wind, you gain some appreciation for how significant this generation source has become in many markets. The chart above shows the 12 states that have the deepest penetration of wind power.
Clearly within these states, and more significantly within regions that encompass many of these states, the penetration of wind power is more than significant enough to alter grid management significantly.
The problems faced by wind power have revolved around the variability - or inconsistency - of the power source. This has led to negative pricing and wind curtailment in any region where there is high wind power penetration; both of these conditions warrant further explanation:
Negative pricing: Negative pricing is quite rare outside of the electricity market, so many readers might be skeptical initially. (I welcome further discussion in the comments section). Selling at a loss is certainly not unheard of, but here prices as low as negative $500+/MWh are possible. Negative pricing means that the power company pays to get rid of the energy. Some of this is indicative of line losses and congestion costs, but often the actual transaction price is below $0.
Power companies are continually making decisions on the extent to which their fossil power generation is tamped down versus selling excess energy to their neighbors at prices that may not be profitable. As more and more wind energy is brought online in any region, preferential treatment of local wind power will ensure that local markets would wish to use some of their wind energy, which means that more energy is now supplied in a market that hasn't seen any change in demand (in most markets of note, North American Electric Reliability Corporation [NERC] de-rates the capacity of wind farms when considering the minimum capacity guidelines, so >90% of the capacity of most wind turbines don't count towards the minimum capacity requirements. Due to its unreliability, wind is usually "extra" energy capacity on the grid). Most grids have less than 10 minutes worth of battery storage capacity, so there's no realistic ability to store this energy. It MUST be sold as it's produced, which means that supply and demand fundamentals don't stop affecting prices at production costs, nor do they stop influencing prices at $0/MWh.
ERCOT is famous for negative pricing, with literally weeks of continuous negative pricing in 2009/2010. But ERCOT West has severe transmission difficulties, and many of the problems were reduced by the completion of a 175 MW transmission line connecting ERCOT West to ERCOT South.
I prefer to use MISO data, because there's less caveats attached. The region of the Minnesota Hub in MISO has no transmission difficulties, is well connected to the other 4 hubs of MISO as well as connected to PJM and SPP - easily able to distribute power among more than half the country. Also, despite the fact that Texas has a multi-billionaire spokesman, looking at the Minnesota hub is far more instructive as to how the grid reacts to deep penetration of wind energy. (Using the chart above, compare the penetration of wind in TX to that of ND, SD, MN, and IA - all four of which trade energy through the Minnesota hub).
To get an indication of how much of a problem negative pricing can be, the following chart illustrates the number of hours where the price averaged below $0 (also shown is the number of hours where the average price was less than $10/MWh) traded over the Minnesota hub in the Midwest ISO (MISO).
The RT market in MISO is traded in 5-minute intervals, but the data is only recorded as average hourly LMP. The chart here lists only the periods in which the average price for the entire hour is negative. The lowest average negative price recorded within that time frame was less than (-$300)/MWh.
Clearly, when low or negative pricing occurs, the best course for the electric power providers is to reduce the production of energy. That's common sense.
For nuclear and supercritical coal plants, the ramp/tamp times are measured in hours… the only way to tamp more quickly is to vent steam, and the inefficiency involved in quickly ramping a baseload plant renders the idea of load balancing with baseload power to be flat-out absurd. NG peakers can easily be ramped up or down in 5-10 minutes, and NG combined cycle gas turbines [CCGT's] can be ramped up or down in 15-20 minutes.
If there's no natural gas that can be tamped back, then the power companies will pitch the blades out of the wind so less power is converted. It's far cheaper to just lose - "curtail" - the additional wind power than it would be to try to inefficiently tamp back coal plants only to have to inefficiently ramp them back up again when the wind lulls.
Clearly, in order for wind power to be effectively utilized, there has to be a dispatchable "balance power", or else a large portion of the wind power will just be curtailed off. This means that the calculation of how much coal vs gas is used changes based on the penetration of wind power in the grid.
An EXTREMELY simplified hypothetical example: If natural gas power is 3 times as expensive as coal power, the coal power might be tamped back any night where the projections show 2/3rd or more of the reduced baseload power would be supplied by wind. If natural gas power is twice as expensive as coal power, then the baseload power might be tamped back to whatever extent is projected will be replaced by 50% or higher wind, and far less wind would then need to be curtailed.
The wind industry should re-evaluate its friends
There have been strong political ties between the wind and solar industry lobbies, so the rhetoric often has been confused on this issue - solar power directly competes with natural gas peakers, and many renewable advocates believe the same is true of wind power. However, wind power performs better and becomes a better bargain whenever natural gas prices are cheap.
In 2010, when the price of natural gas was over $4/mcf, Iowa wind curtailment soared as high as 149 GWh (1/6th of total generation was curtailed) in a single month as penetration breached 20%. In 2012, however, even when penetration exceeded 30%, there was only ~13 GWh of curtailment, because at $2/mcf the price of natural gas power is so low that there's no economic risk to switching more power from coal to natural gas.
Supplies are falling, demand will be higher
The price of natural gas fell sharply beginning last autumn, mostly due to an extraordinarily warm autumn, winter, and spring. The power companies responded by eagerly switching from coal to natural gas. This allowed an extreme reduction in the instance of curtailment and negative pricing. However, it is clear that prices cannot remain this low. The following chart, which readers who are familiar with my earlier work have seen before, shows the industry reaction to plummeting prices:
There is a lag between decreased drilling activity and decreased production, which in this case was exaggerated by a substantial number of "spudded" wells that had not been stimulated to begin production... But there obviously must be a point when a prolonged decrease in drilling would result in decreased production.
I had predicted that YOY production would slip to negative by the end of October, and then begin a gradual drop over the next 12 months. As of October 17, the YOY dry production is only 0.6% above that which was seen last year.
But total U.S. supplies are already dropping, as imports have fallen dramatically.
Quantifying the growth
NG-sourced electricity generation has increased dramatically this year: the first 7 months have seen roughly 30% more NG-sourced electricity than the same 7 month span in 2011, resulting in ~1.2 Tcf greater YOY demand for natural gas. Much of this was obviously fueled by lower gas prices, but as we head into a winter which should be quite cooler then last year, and prolonged and sustained reduction in production; it's important to realize that there are more economic incentives to switch from coal to NG than there are to switch back. NG prices are going to have to raise quite a bit before a significant change in NG power penetration occurs.
The recent recovery in NG prices has not run its course. There may be a small dip or correction when NG storage breaks a new record, but that really was a story that was fully understood months ago. There is more than enough excess capacity to handle continued injections for the next month, and as winter begins we should see record rates for withdrawals, partly because the growing wind industry requires excess balance power to allow variable wind production - which means several times as much NG or hydropower must be produced in order to allow for wind production (hydropower has dropped dramatically over the past year, and will continue to do so until the water levels recover from the nationwide drought).
I'm projecting that prices will exceed $5/mcf by April, and I'm expecting that storage levels will drop below 1.4 Tcf by April.
I'll delve into exactly how we should expect this relationship between wind and natural gas to evolve over the coming decade with my next article.
Additional disclosure: I am part of a team developing a market viable grid integration technology (WindFuels), so the difficulties faced by wind integration are of concern only in terms of future market potential of WindFuels. Near term, we anticipate WindFuels will be a very large consumer of NG, so long-term NG trends are also of interest.