Friday, January 25, 2013

Where's the real bottleneck for natural gas? Distribution.

As a scientist and an unabashed nerd, I love data. Particularly, I love it when ready access to data reveals things that are surprising in the face of conventional wisdom.

Graph of wholesale electricity and natural gas prices from ISO-NEAn interesting case comes up with this year's colder winter in New England. (Thankfully, I'm located in damp, icy East Tennessee, where even the threat of ice and snow manages to bring civilization to a grinding halt.) In particular, Meredith Angwin noted an interesting press release by ISO-NE (the grid operator for the northeastern United States), ostensibly pointing to record-low wholesale electricity prices, but containing something more interesting buried beneath - a slow but quite noticeable creep of wholesale natural gas prices upward to the range of $6/MMBtu (1 MMBtu = 1 million British Thermal Units).
At first glance, this seemed a bit surprising to me, given that while spot prices for natural gas have nudged a bit upward, they're still hovering well under the $4/MMBtu mark; in other words, they don't seem to be going anywhere fast. What really began to catch my eye however was the cyclic behavior of gas prices in the chart from ISO-NE, something which doesn't show up in spot prices from Henry Hub (which generally sets the market spot price for natural gas in the U.S.) 

Natural gas spot, retail, and electricity pricesLooking to verify the trend, I dug a little further around EIA's website. While unfortunately their data on "citygate" prices are a few months behind, the regular periodicity in the citygate price was likewise not there - instead, taken as an average across the U.S., spot, wholesale, and electricity costs derived from natural gas tend to have a strong correlation. Yet here we see above in the data from ISO-NE that prices clearly are deviating substantially from spot prices - what gives?

Natural gas prices spike in the NE corridorIt turns out in fact that the culprit is in distribution. A look around EIA's website brought me to this interesting report, which notes that supply bottlenecks in U.S. northeast for natural gas are expected to produce significant variances in energy prices from the rest of the U.S., and in particular from Henry Hub prices.

In essence, despite a relatively abundant supply of natural gas at the wellhead due to the proliferation of wells seeking to exploit unconventional resources, one thing the laws of physics haven't changed for is the capacity of distribution infrastructure - in other words, pipeline capacity. Natural gas doesn't really care where its end destination is - be it for electricity or home heating. Which means a cold winter can easily drive up demand and stress pipeline capacities - precisely what is occurring, according to the EIA report: utilization rates at the Algonquin compressor station have averaged around 86% for the months of November through December 2012.

Capacity versus price comparison for Algonquin compressorBasic economics can predict what happens next. Because natural gas is generally shipped across pipelines as a compressed gas, due to frictional losses across the pipeline, it must be repressurized at terminals across the pipeline network. The higher demand for gas goes, the closer to maximum capacity these terminals reach. And, as EIA data helpfully shows, the closer utilization reaches to 100%, the larger prices begin to "spread" from spot prices at Henry Hub.

In other words, while the commodity price of gas may indeed be cheap, the wholesale cost to utilities can be an entirely regional phenomenon. This is especially true in the blustery cold of New England winter, where demand is especially cyclic.

Citygate prices in the Northeast versus U.S. average
I compiled together citygate prices for natural gas across the Northeast and compared them to U.S. average citygate prices, and the effect is quite clear - states in the Northeast pay on average an appreciable premium on wholesale prices over the U.S. average wholesale, precisely due to these types of bottlenecks, particularly during times of peak demand (i.e., cold winters). Obviously, pipeline capacity has been steadily increasing in response to demand for gas, however the real issue will inevitably be cyclical "spikes" due to competing uses of gas as a heating source.

Why go to all of this trouble to look at trends in gas consumption (especially when I'm not an energy economist by trade)? Namely because it upends some recent "conventional wisdom" about energy; namely, gas prices are still a regional phenomenon.  While there are some places where supply is not as constrained by distribution capacity (or driven by cyclic consumption) and it thus makes perfect sense to look at natural gas a short-term replacement electricity source (particularly for coal), the Northeast is a shining example where this is definitely not the case. Which in turn makes efforts to shut down stable and relatively low-cost (not to mention carbon-free) baseload sources like Vermont Yankee (and Indian Point) all the more insane - despite the claims of proponents to this end, the balance of electricity from these plants would not be made up from renewables but rather almost certainly come from natural gas. To which again, low spot prices for natural gas are in fact a remarkably poor indicator - especially if one considers what adding an additional consumption driver would look like in terms of utilization "bottlenecks" during times of peak demand (i.e., winter).

I still remain unconvinced of the argument that wholesale natural gas prices are set to explode (and so stand by my bet with Rod Adams) - namely because of the fact that an abundance of known, recoverable supplies makes economics of the raw commodity relatively self-correcting. (In other words, as gas prices go up, so too will the number of wells, pushing prices back down to an equilibrium). However, revolutions in the recovery of gas from novel geology hasn't changed the fundamental physics of pipelines - which in turn may be the real constraint to natural gas growth as an electricity source, at least on a regional scale.

Update: Via Twitter, Rod Adams points to an EIA alert from yesterday about natural gas prices and capacity in the Northeast, given the recent cold snap. Current local spot prices at Algonquin and Transco Z 6 NY (a New York-based distribution hub) are around ~$30/MMBtu - almost 10 times the Henry Hub spot price, with utilization factors reported to be at over 85% capacity. As Rod puts it, "Winter happens."


  1. @Steve

    Do you think that gas marketers have never noticed this effect? Do you honestly believe that they think it is "insane" that they can occasionally sell every therm they can deliver at prices WAY above the price that they paid to purchase the gas?

    No recent news has disabused me of the theory that the major pushers behind the efforts to shut down Northeast nuclear plants are people who make a ton of money by selling natural gas and distillate fuel at prices way above their cost of production. They have rented (or purchased) some very effective politicians to carry their water. The thing I still fail to understand is why there are some otherwise reasonable people who seem to care about the cleanliness of the environment who are going along with the pro fossil fuel effort to destroy electricity production facilities that run on uranium instead of hydrocarbons.

    Rod Adams
    Publisher, Atomic Insights

    1. @Rod: I have little doubt natural gas marketers have a significant financial interest in creating new demand by knocking out competitors by exploiting the force of the government. This is not exactly a surprising claim. I do think you're missing the basic economics of supply and demand - the highest marginal price per unit is not necessarily the profit-maximizing one, namely because it depends upon the elasticity of demand - but whatever. A feature to notice is that this situation only arises when pipeline capacity hits near-full utilization - in other words, it's a price signal for to build more pipelines.

      But in any case, I'm not trying to convince natural gas marketers: I'm making the case that on a pure economics standpoint, dumping pre-existing nuclear for natural gas makes zero sense in the Northeast, something which may not be completely obvious if one only pays attention to the spot price of natural gas. This is the whole point, and one which I'd hoped you of all folks would appreciate - namely that the price of gas is much more controlled by regional variation than is commonly reported, at least in the Northeast. This, in my mind, is a far stronger argument against gas than hypothetical future arguments about price - namely because the data is here in the present and pretty unambiguous.

  2. @Steve

    BTW - I just want to remind you of the terms of our wager - my contention is that natural gas being sold to US electricity generating customers will be trading for more than $10 per MMBTU for at least one month before the end of 2014.

    The specific table I am intending to use as the metric is located at; the row I am referring to is the bottom row labeled "Electric Power Price".

    That particular metric is calculated based on averaging out the prices that electric power generators pay. When there are delivery constraints and the generating companies have decided against purchasing sufficient long term contracts, the elevated spot market prices get reported.

    Therefore, the analysis underlying my bet is not about the "raw commodity" price or reflective of my estimates about the available resources in the ground.

    1. @Rod: Those were not the original terms. I will quote your exact words for reference:

      "@Steve – here is a public wager. By the end of 2014 there will be at least one month in which the price of natural gas at Henry Hub will exceed $10 per MMBTU.

      The bet is $50. Are we on?"

      You stated further in the same thread:

      "@Joel – the bet is for one month with a reported average price at Henry Hub of $10 per MMBTU by the end of 2014.

      That level will be enough to destroy some demand, so I am not sure how long the level will last. I could have slanted the bet more in my favor by picking New York City gate as the delivery point or by stretching the period until the end of first quarter of 2015, but I’m still pretty confident.

      It would be wonderful if I was wrong, but I don’t think that’s true."

      The price at Henry Hub *is* the commodity price, and the one you originally stated in your bet. You may have intended to bet about the retail electricity cost (an obviously more favorable bet to you), but that is not the assertion you originally made, nor is the average electric cost in any way equivalent to the Henry Hub price. And that's the whole point of the post, which I'm surprised that you seem to have missed - I have no dispute that citygate prices in a congested distribution region like New York will potentially spike. However, my bet was that the raw commodity price would stay relatively flat. Your original assertion was that they were not.

    2. Steve:

      We are still on.

      No excuses, but let me explain the confusion. I tried to find a post with the terms of the original wager. I thought I had written a subject focused blog on the topic. The best I could find was the following:

      That post includes the following quote mentioning our bet:

      "I have a public bet with one of my readers that natural gas being sold to US electricity generating customers will be trading for more than $10 per MMBTU for at least one month before the end of 2014. Would you like to make a similar wager?"

      I should have written a special post or at least put the original bet in a more searchable form.

      The market still provides me with some confidence that I have a chance to win the wager, just a lower chance than I thought.