Git-r-done! It is Time to Break Down the Barriers Between Natural Gas and Electric Power

Contributor: Dr. Vince Kaminski, Rice University.  When I was working for Enron in the 1990s one of the mantras of the company was that eventually all the network industries would become like the natural gas business. After all, we are talking about moving hydrocarbon molecules along pipes: electrons along transmission lines, bits and bytes along fiber optic cables, right? Of course, some of these claims were not only patently absurd but also dangerous.   For example, the molecules of CH4 are perfectly fungible. Try to rearrange the zeros and ones behind this post and you will get something much different (it’s even possible it will be something much better). The perception that the company had all the required skills to expand into new and potentially very profitable fields seduced the management to launch a number of risky ventures with less than glorious results.

However, from the perspective of time, one has to recognize that Enron got one thing right: the natural gas and power industries are joined at the hip and it is in their mutual interest to improve the quality of this connection. Given that natural gas is an important fuel source for the US power plants, and most additions to our generation capacity over the next ten years will be gas-fired units, it is important to identify and address all the existing and potential “seams problems,” the impediments to smooth and uninterrupted interactions between the two industries. Given the current abundance of natural gas and low prices, a recent NERC study, “A Primer of the Natural Gas and Electric Power Interdependency in the United States,” December 2011, should be a mandatory reading for any executive in an E&P and or a pipeline company.

The study identifies a number of potential challenges to coordination of technological processes in the two industries and sees a potential for amplification of “the bulk power system’s exposure to interruptions in natural gas fuel supply and delivery.” The challenges arise from the unique nature of electric loads, which translate into many headaches for the pipeline operators: high point loads, high pressure loads, large variation loads, and non-ratable takes. The issues, raised by the study, can be summarized as follows: [1]

  • High Point Loads: Relative to other customers, electric units represents very large point loads.
  • High Pressure Loads: Largely because of improvements in generation technology (e.g., the aero derivative combustion turbines) the pressure requirements for electric loads are much greater than those for other consumers.
  • Large Variation Loads: Historically, gas-fired generation has been at the margin, used primarily to meet intermediate and peaking electricity requirements.  Daily load requirements can be subject to significant variation, as a result of weather events or unplanned outages for other units.
  • Non-ratable takes: Most pipelines are designed to provide relatively uniform service over a 24-hour period. However, there is a limit on the amount of hourly flexibility that a pipeline can deliver (i.e., burning 24 hours worth of gas with an 8 hour period). Furthermore, pipeline flexibility is greatly reduced should all firm customers take their full entitlement to service.

These challenges have been with the market for decades, and the gas and power industry have found workarounds, compromises and convoluted processes to make things work.  One of these is “No Notice” service, a legacy of the bundled pipeline services world of 20 years ago in which the pipeline gives a power plant the right to swing up and down on its system supply as needed, with “no notice” and no scheduling.  It is an expensive service that has the advantage of requiring little coordination between the pipeline and the power generator.  It has the disadvantage of high cost, which can work to limit the generation of power using gas as a fuel by pushing its marginal cost higher, and thus up the generator’s merit order curve.

In a perfect world, scheduling of natural gas supplies into power plants would be a seamless process, similar to how the two industries do business within their own domains.  But it is not a perfect world.  One of the most frustrating and seemingly intractable problems is simply the definition of “DAY”.

In the power world, a Day is from midnight-to-midnight, transitioning at the low demand, midnight drop off point.   Operational planning wraps up about 6pm the prior evening, and real time economic dispatch makes adjustments during the Electric Day.  In contrast, gas which is to flow “next day” trades in the morning prior to the day of flow.  Nominations (schedules) are in by 11:30am, and the Gas Day runs between 9am-to-9am central time the next day (Houston time, of course).  Thus the gas day transitions exactly when power demand is ramping up. 

According to NAESB (North American Energy Standards Board) rules, there are three more “bites at the apple” when natural gas flow schedules can be adjusted.  However, these adjusted schedules are typically much more susceptible to cuts and allocations by the pipeline to stay within operating parameters (See Rusty’s post yesterday on Operational Flow Orders, or OFOs.). 

Over the years, the power and gas industries have made a few attempts to reconcile these issues, to no avail.  Until now, this was disappointing but tolerable.  But the shale phenomenon has changed the balance of power in the energy world.  It should no longer be tolerable. 

The NERC study is a great starting point, and recommends improved information sharing, coordination of business processes, sharing of information in real time, and greater reliance on natural gas storage as a buffer between the two grids.  Implementation of these recommendations should be the number one priority for a joint power-gas industry task force.   If natural gas is to reach its full potential as an abundant, affordable, clean and home-grown energy source, it is imperative that the natural gas and power industries take on this challenge.      Now is the time to “git-r-done.” 

[1] NERC Study Page 84

Rusty Braziel contributed to this post (See if you can figure out what parts).