2002-03 gas storage season presents a supply conundrum

June 9, 2003
Projections of growing US natural gas demand and concerns about the ability of North American producers to meet that demand raise doubts about industry's ability to fill gas storage this summer while at the same time meeting the needs of electric generators, industrial users, and other requirements.

Projections of growing US natural gas demand and concerns about the ability of North American producers to meet that demand raise doubts about industry's ability to fill gas storage this summer while at the same time meeting the needs of electric generators, industrial users, and other requirements.

The potential gap between forecast US baseload natural gas requirements and production is best served by new production and incremental pipeline capacity.

At the same time, the peak day, peak month, and seasonal load requirements of the U.S market are growing even faster than the base load demand, and this is increasing the importance of the various peaking services, primarily underground storage and LNG peaking facilities.

International Gas Consulting Inc., Houston, believes the infrastructure in place has served demand efficiently and effectively, but the margin of confidence to be able continually to perform has become slim.

To meet these challenges, the US gas market must evolve to support an increase of storage capacity close to the demand centers and an improvement to efficiency focusing on storage deliverability.

The location and flexibility of gas delivery service comprise the key source of value for peaking services today.

The ongoing need for market efficiencies will force a re-engineering of existing infrastructure as a part of a gas market of the future.

At the same time, the decline of the mega-marketers in the energy business has fueled a return to assets, as opposed to contracts, as the guarantor of liquidity. Summer 2003 may prove to be a watershed for developers of new storage as interest in these hard assets will encourage new interest in efficient new storage services offered close to city gates.

The need to refill storage throughout the year will also become increasingly important as the wellhead and imported LNG supplies look more like a "just in time" supply system rather than the "gas supply bubble" of the 1980s and early 1990s.

A look back

The 2002-03 natural gas storage year has been dramatic in an industry traditionally not known for drama. Industry press included "breathless" headlines sounding alarm over record high inventories in April at the beginning of the injection season.

As the year progressed, a shift occurred to equally disconcerting reports of record storage withdrawals, and early spring news voiced concerns about the low end-of-season inventory in February and March.

Now, public concern has shifted to the issue of storage fills for summer 2003 with predictions of a catastrophic shortfall of storage gas in the next withdrawal season.

Among other issues, this was the first storage cycle to occur after reporting responsibilities had been shifted from the private sector to the a governmental agency. The American Gas Association had been collecting and reporting weekly storage inventory data since 1994. In May 2002, AGA turned over that responsibility to the US Energy Information Administration of the Department of Energy.

The industry waited, wondering if the government agency would be able to provide reliable and timely information. After early reporting "hiccups," EIA numbers now appear reliable and are used by industry with similar confidence to that given AGA's values.

EIA's Weekly Storage Report stated national storage inventory levels on Apr. 1, 2002, at 1,415 bcf. This was the highest storage inventory experienced at the beginning of an injection season since the AGA began reporting weekly gas storage inventories in 1994. The press declared that there was an 800 bcf surplus of natural gas in storage inventories.

Injections proceeded throughout the season from this relatively high starting point. Week after week, reports detailed the continuing surplus inventory levels. Steady injections totaling 1,757 bcf during the injection season led to a Nov. 1, 2002, inventory of 3,172 bcf. This inventory at the beginning of the 2002–03 winter season was only slightly higher than the previous year.

Then by the winter and early spring, storage withdrawals proceeded rapidly. By mid April 2003, just beyond the end of the traditional withdrawal season, 2,549 bcf had been withdrawn from storage. This left only 623 bcf of gas in inventory and set a new record for gas delivered from storage in a single season. The reported surplus had become a 1-tcf deficit.

Click here to enlarge image

The terms "surplus" and "deficit" suggest that the stated inventory levels for most of the season were somehow wrong, that there was either too much or too little natural gas in storage. Fig. 1 shows the cycle of weekly natural gas storage inventories in US storage facilities for each of the past 9 years.

The total maximum inventory tends to cluster around a fairly consistent number. In 6 of the last 9 years, the maximum inventory level was within 120 bcf of 3,070 bcf. This consistency is especially remarkable considering that it is the product of unpredictable weather and the independent decisions of hundreds of individual gas supply and storage managers.

The date that the maximum reported inventory occurs varies year to year, depending primarily on regional weather patterns. If winter starts early in areas that depend on storage, the net activity switches from injection to withdrawal earlier. If winter is late, net injections continue a little longer.

The maximum inventory value tends to grow a bit each year, along with increasing total gas demand and the availability of new storage facilities.

A greater variation year-to-year occurs at the end of each season and, of course, the corresponding beginning of the next season. Since each winter is different in both severity and the timing of cold weather, the pattern of withdrawals varies.

Also each storage manager has a different set of objectives regarding their own individual ending inventory that will further contribute to the variation. Some are subject to restrictive tariffs that require that they withdrawal gas to a level below a specified maximum ending inventory.

Some may be encouraged by their management or regulators to minimize the inventory carried forward into the next year. Others may see an opportunity to maintain relatively high ending inventories if they can roll their futures market positions forward and capture additional margins. Wide variations in weather, prices, and motivations naturally result in substantial changes in year-over-year inventory withdrawals.

The relatively low inventory, 623 bcf, at the end of this year's winter season should not cause alarm in and of itself; lower inventories have been experienced. Since 1994, there have been 6 weeks that gas storage inventories fell below this level, although 5 of these 6 weeks were in the winter of 1995–96. Storage inventories hit the lowest level of the past 9 years in mid-April 1996, bottoming out at 546 bcf, 77 bcf below the lowest level in 2003.

The real issue is the 2,549 bcf withdrawn from storage in the 2002-03 winter season. This is the net change in storage inventory from peak inventory volumes at the beginning of the withdrawal season to the minimum inventory levels at the end of the withdrawal season. Fig. 2 illustrates that this unusually large withdrawal is the largest net storage inventory change in 9 years.

In Northeast and Midwest US regions, the vast majority of the storage facilities were designed and built for single cycle, seasonal withdrawal and injection, with little flexibility in deliverabilities. The older, depleted reservoir storage facilities of the Northeast and the aquifer and reservoir facilities of the Midwest are not necessarily capable of providing (or engineered to provide) a quick response to fluctuating needs.

Not considered in this analysis of seasonal net withdrawals is the effect of short-term storage cycling, i.e. daily injections and withdrawals as is possible at salt cavern storage facilities in the Southeast and in Texas and Louisiana. This daily activity is not necessarily reflected in the EIA reports.

Notably, market demand was for the most part satisfied throughout the 2002-03 winter. Despite some near misses, the system worked as it should have. The resulting low level of storage inventories presents numerous challenges. Most importantly, it raises the questions: Will there be sufficient gas in storage to meet next winter's gas demand? Will existing storage facilities and flowing gas supplies on pipelines be capable of meeting evolving requirements in future gas demand?

The US storage industry is now asking: Just what type of storage is really necessary in the next decade, considering the dramatic changes in the midstream energy industry? This winter's activities shed some light on these questions.

Unprecedented withdrawals

The variability that affected the 2002-03 storage season began with the after-effects of the winter of 2001-02.

According to the National Oceanic and Atmospheric Administration, the 2001-02 winter season nationwide was 14% warmer than normal, based on calculations of 1961-90 data and national average heating-degree days (weighted by gas home customers). In addition, the forward price curve at the time showed a modest but significant growth in prices.

The warm 2001-02 weather meant that there was no need to draw down storage inventories to low levels, and the price curve in place at that time provided little economic incentive to do so. These two factors contributed to record high inventories at the beginning of the 2002-03 storage injection season.

The fact that the market remained in "contango" throughout the injection season provided an economic incentive to continue injections at close to normal rates regardless of the starting inventory level. (A market in which futures prices are progressively higher in later delivery months is referred to as a contango market.)

Independent of the price signals, individual utility supply managers had the traditional obligation to have sufficient gas in storage to ensure winter supply for their customers.

Together these factors resulted in an injection pattern that kept inventories historically but not extraordinarily high, through the entire injection season.

As further testament to the ability of existing processes and infrastructure adequately to fill storage inventories, US Gulf Coast gas production was reduced by several major storms during summer 2002; storms Isidore and Lili resulted in significant gulf production curtailments.

Despite these supply disruptions, summer storage injections continued at an adequate pace.

Nationwide, winter 2002-03 averaged slightly warmer than normal, primarily because of warm weather in the western and producing regions. (The western region includes Mountain and Pacific states; producing region includes western south central states.)

The area most dependent on storage, the eastern region including New England, Middle Atlantic, and eastern north central, was slightly colder than normal.

This past winter season was one of only three winters over the past 9 years that have been colder than normal.

Click here to enlarge image

As Fig. 2 shows, gas demand resulting from cold weather was not the only influence in the large storage withdrawals. Winter 2000-01 was colder than the most recent winter, but 2000-01 had significantly less storage withdrawal.

The graph shows a consistent correlation between cold weather and storage withdraws 1994-98, but since the late 1990s, there have clearly been other factors besides the weather that have significantly affected storage withdrawals.

Click here to enlarge image

One possible explanation is price. In fact, however, natural gas commodity prices have had only a marginal effect on storage injections and withdrawals. Fig. 3 shows historic storage activity relative to Henry Hub (La.) monthly prices.

Net injections typically approach 100 bcf/week in summer regardless of gas price increases.

The obligation to serve and reliability of supply remain key drivers in the need for local distribution companies to have storage full at the beginning of the winter season, and the timing of injections can only be modified on a short-term basis if supply costs are high.

For many LDCs, the absolute price of gas has very little impact on the decision to put gas in storage. Over a season, the cost of the storage gas in inventory is averaged and managed prudently to mitigate market volatility and reduce the impact of price spikes during the winter heating season. Furthermore, withdrawals increase to 200-250 bcf/week in the winter even if high spot prices encourage more aggressive withdrawals.

In general, marketers use storage activity as a price signal. Day-to-day, storage is a balancing item between production and demand. Gas that is not consumed is injected; gas that is required but not currently produced is withdrawn. Net injections and withdrawals are seen as a proxy for real time information on consumption and production.

Marketers use storage data to adjust their perceptions about the supply-demand balance. While there is a complex interaction, storage activity is more likely to drive price than be driven by it.

The lack of market liquidity caused by the recent demise of the energy traders and supply aggregators has had a tangible effect on 2002-03 storage activity.

Before to the recent mass exodus from gas marketing and trading, supply aggregators offered "value-added" gas supply services, such as emergency back-up, swing, and balancing.

Typically, a marketer would offer these services in multiple markets, spreading the costs and use of storage capacity over all sales volumes and creating efficiencies unavailable to a single end-user in a solitary weather pattern. In many respects, gas markets have now regressed to favoring more conservative asset-based, low risk transactions.

Fewer players mean fewer options for supply and the traditional answer when supply is tight is to "pull from storage." Fewer and smaller marketers mean less opportunity to take advantage of non-coincident peaks, which, in turn, also translates into "pull from storage."

Overall, gas-market uncertainty also contributed to hard storage pulls. The return of near-normal temperatures had a negative psychological effect after several warm winters.

Prolonged cold weather stressed the distribution systems of natural gas, propane, and other heating fuels. Well freeze-offs in the MidContinent and Permian Basin in January 2003 caused supply interruptions and aggravated doubts fueled by fears of diminishing gas production.

Summer fill

After the successful performance of the storage through the somewhat normal winter 2002-03, the uncertainties of the next season loom.

Gas demand is expected to increase, while gas supplies are at best flat. Will the existing infrastructure be sufficient to meet growing demand? What if we experience both increased demand and a severe winter?

Gas supply and wellhead deliverability concerns are undeniable. In an Apr. 10, 2003, policy-analysis issue, the AGA reported that the US reserve base showed no growth in 2002, the first time that has occurred in 4 years. At yearend 2002, reserves totaled slightly less than that 183.5 tcf reported at yearend 2001.

Gas demand is growing more quickly than new sources of production are being found. Increased rig counts and higher production rates have in many areas only increased the rate of depletion, not added new gas supplies. Higher gas prices should encourage additional drilling, but the timing of first bringing the supplies online and then to market remains a concern.

Undoubtedly, one factor in the heavy withdrawal rates experienced this past season was the need to supplement production.

LNG will be an increasingly important component of the US supply portfolio. Although all of the LNG facilities in the Midwest and many in the Northeast are captive to a specific LDC, LNG-receiving terminals at Boston, Cove Point, Md., Elba Island, Ga., and Lake Charles are connected to the interstate pipeline grid. This allows these facilities to baseload gas deliveries (in addition to offering peak shaving).

These LNG receiving terminals are unlikely to reduce the seasonal use of storage and may actually require additional underground storage in place of surface tanks as through-put increases at the active terminals.

The demand side of the equation is also a concern. LDCs have residential customers who need gas for heat regardless of price; a severe winter could test the capability of the existing infrastructure.

Summer injections must be made, while increasingly limited gas supplies are being called on to meet counter-seasonal gas-fired electric generation loads. To date, balancing these two competing high-priority requirements has been accomplished effectively.

The issue is further complicated by the full requirements of the electric-generation industry. Most US electric utilities are summer peaking utilities. That means that they have their peak requirements when the air conditioning load is highest and the demand for gas to serve LDC customers is lowest.

While this means that electric generators may compete with LDCs trying to inject gas into storage, they are counter-cyclical to the primary LDC load profile. Many electric generators, however, especially in northern climates, have a secondary peak in the winter to serve their own heating-demand customers.

That suggests that as gas becomes a larger percentage of the power-generation fuel mix, electric generators will have an increasing need to procure gas supplies in the winter. This will contribute to the pull from storage.

Regardless of the season, gas-fired generators have the potential to decrease storage capacity available to other end-users if storage is contracted to provide load-following gas supply for peak intervals of power generation.

Baseload gas demand may be reduced if gas prices stabilize at levels exceeding $5/Mcf. Certain industrial end-users, such as petrochemical plants and fertilizer producers, are highly sensitive to increasing gas prices because natural gas is a major component of their production costs.

Fuel-switching is an option available to a small percentage of dual-fuel generating plants and industrial end-users, but this reduction is likely to be more than offset by increased demand from anticipated additions of new gas-fired generation.

Thus, storage use could increase due to increased fluctuations in gas demand (i.e., a demand curve with more peaks), even if there is a small decrease in baseload demand in the industrial sector.

Substantial public debate has centered on the need to refill storage to the 3 tcf. IGC's analysis indicates that the current LDC market supports closer to 2.5 or 2.6 tcf of storage capacity, with traders and marketers utilizing the remaining storage capacity.

Thus, if the forward price curve fails to provide financially attractive injections, the marginal storage volumes will not be required. This will almost certainly cause strong pricing during peak winter demand for days or weeks.

A reduced storage fill will not likely cause a catastrophic failure of the nation's gas supply, but it will likely afford holders of firm pipeline capacity rewards not often seen in the daily gas markets.

Click here to enlarge image

The authors
Debra A. Caperton has more than 18 years of energy industry management experience. She has contributed in both a technical and commercial capacity on storage projects at IGC for 3 years. She has worked in areasof increasing responsibility on the geological and geophysical staff at TransGulf Energy LLC, Tenneco Energy Corp., and Texas Eastern Exploration Co. At Tenneco Energy, she served as manager the regulatory affairs. Caperton holds BS in geology from the University of Oklahoma, completed Rice University's Management Development Series, and attended the master of business management program program at Houston Baptist University.

Click here to enlarge image

H. C. "Rusty" Cates is the vice-president of International Gas Consulting Inc. and has more than 20 years of energy industry experience with, among others, Phillips Petroleum Co. and Texaco Inc. Formerly, he chaired the Natural Gas Supply Association's Federal Regulatory Affairs Committee. Cates holds a BA in history from Yale University and an MBA from the University of Tulsa.

Click here to enlarge image

Kenneth L. Beckman is president of International Gas Consulting Inc. and is a natural gas market analyst and petroleum geologist with more than 20 years of varied domestic natural gas marketing, gas transportation and exchange, and international petroleum exploration and production experience. He has served in various areas with Amoco International Oil Co., Union Texas Petroleum Corp., and Bishop Pipeline Corp. In addition, he worked 3 years as an independent geological consultant. Beckman holds a BA in geology from the University of California at Santa Barbara and an MBA in finance from the University of Houston.

Click here to enlarge image

Keith M. Schultz has more than 14 years of natural gas experience as a leader in the performance of economic analysis, training and development, marketing, and strategic planning. He acquired his experience through increasingly responsible positions with MidCon Corp. Schultz holds a BA in economics from Allegheny College, Meadville, Pa.