Issues, Trends, Technologies
Storage downstream of regasification optimizes LNG’s value
John H. Holcomb
John H. Holcomb, Falcon Gas Storage Co., Houston
Few observers today dispute that LNG will soon play a major role in supplying the needs of the North American natural gas market. As US domestic production and Canadian imports struggle to supply growing demand, LNG’s role will be critical to help ensure that North America has a long-term reliable supply of natural gas.
Since LNG cannot efficiently be stored, however, underground gas storage will be essential to providing LNG with a means to optimize its value by capturing seasonal price arbitrage and serving the growing power-generation market.
Although each terminal has LNG storage capacity on site, LNG cannot efficiently be used to take advantage of seasonal arbitrage opportunities or to serve the daily load profiling requirements of gas-fired electric generation (“GFEG”) consumers because the liquid storage must be vaporized and depleted before each new cargo of LNG arrives. Storing LNG is simply a delivery mechanism and is not nor ever was intended to be a subheadtitute for gas storage.
Ultimately, incremental development of gas storage will be driven by the value it provides to its customers (i.e. LNG suppliers) and its competitiveness with other storage alternatives. Compared to storing LNG, gas storage is both economically and operationally superior.
The key debate will be over which type of gas storage will be optimal for LNG: in high-deliverability multi-cycle (HDMC) gas reservoirs or in salt dome formations. HDMC storage has economic benefits due to its competitive cost structure and ability to capture contango spreads and load-following value because of its high deliverability and multi-cycling capability. (Contango is a market condition in which longer term futures contracts carry a higher price than shorter term contracts. This condition encourages inventory builds if the contango exceeds the costs of carrying natural gas for future delivery.)
Current markets
The US consumes slightly more than 22 tcf/year of natural gas, a rate that is projected to grow at roughly 1.5%/year. By 2020, consumption could reach 28 tcf/year, an incremental 16 bcfd over current demand. US gas production and Canadian gas imports, however, will in aggregate be unable to keep pace with growing demand.
Domestic gas production and Canadian imports may, at best, remain flat and, may, in some industry experts’ scenarios, soon begin an irreversible decline just as crude oil production did in the middle 1970s. Hence, LNG imports will be required to fill the growing gap between demand and supply.
As Fig. 1 shows, LNG imports are expected to grow continuously for the next 15 or more years. By 2010, LNG could account for 10% of total US gas supply. By 2020, it could be 20%.
Demand for natural gas is not constant throughout the year, however. Traditionally gas storage is used to supply the seasonal gas demand by smoothing the deliverability.
As Fig. 2 shows, seasonal gas demand variations are huge. The total swing from summer to winter can be 60 bcfd or more. These seasonal swings are likely to become greater as residential and commercial heating loads continue to grow and more gas-fired power generation is added.
This current summer-winter gap is served by gas storage. Roughly 3.3 tcf of working gas underground storage capacity are throughout the US and Canada. These 385 or so facilities can deliver up to 50 bcfd of withdrawals and inject up to 35 bcfd. Current gas storage can barely handle current seasonal swings and more is needed to serve the growing gap between supply and demand. The market has evolved to the point where gas storage is not only needed to meet winter demand, but also to meet gas-fired electric generation summer load requirements as well.
Today environmental concerns have made natural gas the fuel of choice for GFEG. Gas-fired capacity makes up almost 20% of the total generation mix, up sharply from only 10 years ago.
This has created a new problem for the gas pipeline industry. Pipeline infrastructure was developed to transport natural gas at a constant volume each day (“ratable” flow: gas flows at the same volume every hour of the day). GFEGs typically operate when there is demand, however, since electricity cannot be stored.
Therefore, gas is only consumed by the GFEG when electricity is needed. Peak electricity demand is daily from morning to evening. That means GFEGs are ramping up electricity production in the morning, peaking around late afternoon, then ramping down until late evening. The excess generation is shut down completely at night. This creates non-ratable flow on the pipelines, and gas storage is rapidly becoming the only effective means of balancing this type of flow.
Gas storage value
The simplest method to determine the value of gas storage is to measure the intrinsic value of the forward natural gas price curve. Of course one must net out costs: the change in basis differentials (if any), the cost of storing the gas, and the incremental cost of transporting it into and out of storage.
The value of gas storage is more complex when it includes the extrinsic value of storage. That value is determined by modeling the forward volatility of the price curve. Extrinsic value of storage can add as much as a two to three times to the intrinsic value.
Increased seasonal forward-price volatility has significantly increased the intrinsic value of storage. In part, this is due to changing market dynamics from increased gas-fired electric generation load and from the change in the supply-demand balance, which has shifted in recent years to a “supply short” situation: The “gas bubble” is gone.
Fig. 3 shows how the contango spread values have increased over the past 2 years. The graph shows the actual New York Mercantile Exchange (NYMEX) forward price curve at three distinct times. In less than 24 months, the 4-year average contango spread has increased from $0.98/MMbtu to $2.31/MMbtu, a 135% increase. This intrinsic value can be locked in as long as the storage facility has the physical capability to deliver gas in those months when a hedge is transacted. That’s one reason storage development is hot.
Another significant value driver is serving the intraday gas requirements for the GFEG market. Gas storage to balance its daily gas requirements has become increasingly important and profitable.
Generally speaking, GFEGs consume 90% or more of their gas requirements within 16 hr. Therefore, gas purchased under a normal 24-hr ratable gas North America Energy Standards Board cycle is actually burned during an 8 to 16 hr period. Gas storage allows a GFEG to continue purchasing gas on a ratable basis under NAESB standards, but also inject gas during the night and withdraw gas during the day. Pipeline nominations remain ratable and balanced, while the storage facility accommodates the GFEG’s load profiling.
The value of this service is driven by two primary elements: other storage alternatives and the difference in hourly power prices and daily gas prices. Traditionally, this service was provided by the pipeline companies which offered “synthetic” storage: packing their lines at night and unloading during the day. In the past, this intraday service from the pipelines was heavily discounted or provided at no cost.
But that’s not the case today. As pipeline capacity utilization has increased and more gas-fired power plants have come on line, pipelines are unable to offer this service without charging for it.
The other element is the hourly pricing of power vs. the daily price of gas. Since electricity cannot be stored, gas storage offers an option to the power market, an intraday call or “put option,” as the case may be, for intraday power demand.
If gas storage is used as the intraday call option for the power market, then the option price for the intraday service adds to the replacement price for the gas. The replacement gas being purchased in the forward market drives up the forward price of gas. The option provider, or buyer of gas, is willing to pay a price for the replacement gas equivalent to the forward price plus an amount less than the intraday option value in order to keep some room for profit.
If the market is in contango, the premium for the replacement gas can get very high and, likewise, if the market is in backwardation (the opposite of contango), then the premium is reduced to only the intraday option value.
Comparisons
New customers for gas storage will be LNG suppliers wanting to optimize the value of their gas. Most of the planned or permitted North American terminals are likely to be located along the Gulf Coast where pipeline infrastructure delivery system is well developed and the business environment is industry-friendly. This region is also well suited for gas storage, since the Gulf Coast has an abundance of good geological formations that can be developed for both salt dome and depleted gas reservoir storage.
LNG operational characteristics require that LNG cargoes be unloaded immediately and continuously. LNG tankers are uloaded and turned around as rapidly as possible in order to optimize the use of LNG cargoes.
LNG must be vaporized and delivered into the pipeline grid frequently whether there is demand for the gas or not. LNG onsite storage, on average, has only 5 days of capacity (at maximum sendout capability), making on site storage inadequate for storing gas for long periods of time. In addition, LNG terminals do not have the ability to offer daily load-following service for the GFEG market without having access to gas storage.
The bottom line is that LNG is not suited to provide the same types of services that gas storage offers nor can it capture contango spread values.
Fig. 4 shows projected LNG imports (under a likely scenario) in the Gulf Coast that will arrive within a 60-mile radius of the mouth of the Sabine River. This LNG corridor will be home to an estimated six LNG terminals between now and 2013, adding potential sendout volumes of 13 bcfd. These LNG terminals include CMS Trunkline (existing, at Lake Charles, La), Freeport LNG and Freeport LNG expansion (2008 and 2010, respectively), Cheniere Sabine Pass (2009), ExxonMobil Golden Pass (2010), Sempra Cameron (2011), and Cheniere Creole Trial (2012-13).
By 2010, an additional 4.4 bcfd of gas will be flowing, on average, through this region heading east toward Florida, mid-Atlantic, and Northeast markets. At maximum sendout, this figure doubles to 8.8 bcfd. When Texas exports and Louisiana offshore gas production destined for these same eastern markets are added in, the total flow from this specific region will increase to 12 bcfd (on average) and 16 bcfd during peak sendout.
Some LNG suppliers anticipate using salt dome storage for their storage needs, but salt dome storage will compound the problem. Salt dome storage economics require that users subheadcribe to high inventory turn services, typically from 6 to 12 cycles. That means that the salt dome storage needs to be filled and emptied every 1-2 months in order to capitalize on the higher unit storage cost.
High inventory turns increase the excess supply situation during slack demand periods because both the LNG vaporization and withdrawals from salt dome storage hit the market at the same time, competing for limited markets and limited spare pipeline capacity. Gas withdrawals from salt dome storage could compound the problem even more by adding another 4 to 5 bcfd to the already extraordinarily high volumes (Fig. 4).
The impact on gas prices caused by such potential “basis blowouts” could have dramatic negative effects for the LNG supplier, as BG has found at the Lake Charles LNG terminal when cargoes are offloading. And that’s only one terminal, not six.
Instead, HDMC depleted-reservoir storage located further downstream of the pipeline congestion and closer to the market allows an LNG supplier to retain the value of the contango price spread at better and more competitive storage rates, thereby allowing customers to inject continuously during slack demand and to sell continuously during high demand periods.
HDMC still allows customers to take advantage of the dual peaking summer period and provide intraday load-following service for the GFEG market. Table 1 summarizes the key service objectives, attributes, and pricing of HDMC storage.
Furthermore, the cost component between the two storage options is not insignificant. For example, the savings for reserving 4 bcf of storage capacity amounts to roughly $5 million/year, assuming three turns/year from an HDMC facility vs. eight turns/year service from salt dome storage.
This example does not take into account the potential negative effects of “basis blowouts” that are likely to result from high salt dome storage withdrawals in the same production area where LNG terminals are located. On a unit basis, this cost difference equates to $1.20/MMbtu/year; slightly more than the average historical contango spread.
Considering the higher cost of salt dome storage, HDMC is the better gas storage option for LNG because it provides:
- Excellent deliverability for injections during low-priced shoulder and summer months.
- Good withdrawal capability and the ability to capture contango spread values.
- Multiple-cycle daily operational flexibility to serve load-following requirements for the GFEG market.
- Competitive prices that allow for larger profit potential.
While it is likely that LNG suppliers will contract for some salt dome storage, they should also have HDMC capacity in the mix to help optimize the LNG value.
Author
John H. Holcomb (jholcomb@falcongasstorage.com) is director of marketing and business development at Falcon Gas Storage Co., Houston, one of the largest independently owned developers and operators of HDMC gas storage capacity in the US and the largest privately owned gas storage provider in Texas with more than 22 bcf of working gas storage capacity. Holcomb has more than 20 years of gas marketing, pipeline development, and energy consulting experience, after spending more than 12 years at Tenneco Energy Co. and 5 years at Pace Global Energy Services. Holcomb holds a BS in petroleum engineering from the University of Oklahoma and an MBA in finance from the University of Houston.
Volume 3 Issue 3
Jul 01, 2006
Add RSS Feed