January 17, 2019 | Jack Weixel
In this edition of Get the Point, OPIS PointLogic is going deep cover, but not in the sense that Dr. Dre and Snoop Dogg so eloquently described in Dre’s 1992 hit debut single “Deep Cover”. Instead, we’re going to take a deep dive on the capacity of winter natural gas storage inventories to cover systemic demand increases during the cold-weather withdrawal season.
Lower 48 storage working gas storage capacity has increased by 640 billion cubic feet (Bcf) from 2008 to 2017, or about 20%, according to the US Energy Information Administration (EIA). Over the same time period, Lower 48 natural gas demand has increased by 26 Bcf/d, or 40%, with peak demand months in the winter exceeding 100 Bcf/d on a regular basis. Is the growth in demand relative to storage capacity increasing market risk?
Domestic demand in the Lower 48 US averaged 73.2 Bcf/d in 2018, or 4.3 Bcf/d higher than in 2017. While about half of this increase can be attributed to weather (an increase in residential/commercial demand), over 2.0 Bcf/d was due to increased consumption of natural gas in the electric power and industrial sectors.
Source: OPIS PointLogic Supply Demand Module
Power demand has increased over the past several years for several reasons: an increase in baseload gas-fired power generation units; a tightening coal market with systemic production declines; and a newfound inelasticity to the price of gas, which has muted coal-to-gas switching potential. Therefore, gas is called on to power air conditioners in the summer and electric baseboard heaters in the winter months like never before.
In the industrial sector, we’ve seen much quieter growth. The industrial renaissance in natural gas has been a silent but steady contributor to overall increases in domestic demand. According to OPIS PointLogic modeled data, industrial demand in 2018 set monthly records in every month of 2018, except December.
Export demand is growing, too. Exports to Mexico and LNG feed gas deliveries increased by over 1.6 Bcf/d in 2018, making them the two largest drivers of demand growth on a percentage basis. Exports to Mexico have increased each year since since 2011, while LNG feed gas has seen a meteoric rise since the first commissioning cargoes left Sabine Pass LNG in November 2015.
The market has handled rising demand largely through increases in dry production, which also set records in 2018, up 8.6 Bcf/d over 2017. But as in any market, the co-dependent relationships ebb and flow, depending on market conditions. Several factors bear watching. One is the gas storage. As production has grown, and quicker-to-market shale reserves have materialized, the need for new storage facilities has diminished. In general, producer drill times have decreased, and operator responsiveness to changes in price has increased. As a result, gas storage has stayed relatively constant, especially in the most recent years.
According to the EIA, during 2017 and 2018 only three storage expansion projects have reached the construction phase of development. These are the Crownville Salt Dome Cavern 2 expansion project in Louisiana, the Lewis Creek Field in Colorado and the Mist Storage Expansion Project in Oregon. In total, working capacity at these fields will increase Lower 48 storage capacity by a mere 15 Bcf, or less than 0.4% of current US storage capacity. Another 14 projects have either been shelved, put on hold or are still in preliminary planning stages.
While the EIA estimates that non-coincidental working gas volume storage capacity is roughly 4,300 Bcf, over the past five years the average fill-to requirement determined by the market has only been around 3,800 Bcf. In theory, the market incentivizes companies to fill up storage to its operational capacity (or non-coincidental maximum working gas volume) in the Lower 48 by the end of October each year. This summer, there was a clear lack of concern to fill up storage as end of October 2018 storage inventory reached just 3,189 Bcf, its lowest end of injection season inventory level since 2002.
Then, a quick start to winter in November 2018 put storage inventories front and center briefly. At the same time, there was real fear that a cold winter with elevated systemic demand (such as exports and inelastic power demand) could put inventories at the end of March 2018 at their lowest levels in history. Since that time, however, the lack of December demand has eased these concerns somewhat, but the question remains whether there is adequate storage capacity to cover increases in peak winter demand levels.
Winter Demand in Perspective
In order to fully analyze this concern, OPIS PointLogic has looked at the incremental demand that occurs during the peak winter months of December, January and February for all winters going back to winter 2006/07. While it is evident that average winter demand is increasing, peak winter month demand (December-January-February) is slightly more choppy. Over the past five winters, three of five peak winter periods have been over 100 Bcf, but the winters of 2015/16 and 2016/17 were both in the mid-90 Bcf range. Last December through February saw the highest peak winter month average in history, thanks to the ‘Bomb Cyclone’, which helped to prop up total winter 2017/18 demand to average over 100 Bcf/d for the first time in history.
So, with all this growing demand, is storage capacity adequate?
The role and value of storage is its ability to cover peak demand days and peak demand periods over and above what is considered normal demand. But what is normal demand, and how much incremental demand during peak winter months should storage be able to handle?
The market determines this level throughout the summer as supply, demand and price dictate to what fill level storage should reach by the end of traditional injection season, or October 31. As we witnessed this past summer, certain storage capacity holders are also motivated by the spread between cash and prompt month when deciding how much to inject, particularly in flexible salt storage caverns.
Each year as summer winds down, long-range weather forecasts begin to roll in, influencing the prompt and winter strip prices. The level of supply expected over the winter (based largely on Lower 48 production levels) compared to anticipated winter demand (largely based on weather) impacts the spread between cash prices and the winter strip, and thus the level of injections the market believes is required to cover winter demand. If winter production levels are expected to increase relative to demand, the cash-to-winter-strip spread will tighten (and in the case of this summer, can actually backwardate -- that is, show a declining price over time). Conversely, if winter production levels are not anticipated to be enough to cover winter demand (due to expectations of a cold winter or stagnant production), the cash-to-winter spread will widen.
Anticipated winter demand less anticipated winter supply represents the expectation of what storage should be expected to cover. The actual call on storage during peak winter months of December, January and February is a good measure of whether storage capacity is adequate.
Storage is most heavily utilized in the winter, and it's when a swing in weather, plus gains in systemic demand, could threaten end-of-March inventory levels. What matters is the delta between peak winter demand (December-January-February) versus average winter demand for all winters going back to winter 2006/2007. The graph below shows that incremental demand during peak winter months has not increased over the past 12 winters and has been fairly stable over the past seven.
The graph below represents the starting inventory of winter (as of October 31 in each respective winter) divided by the “call” on storage during the peak winter months of December, January and February (the amount of incremental demand during these peak winter months). This number could be interpreted as the “days of cover” if peak demand persisted over the entirety of winter, ignoring incremental supply available to the market.
As is evident in the graph, “days of cover” is not shifting down with any severe trajectory. Days of cover over the past three years is actually very similar to levels seen from winter 2006/07 (with one exceptional year). Even our forecast winter 2018/19 does not show obscenely high levels of incremental demand (the call) compared to average winter demand.
Another way to look at this relationship is to take starting inventory and divide by actual winter supply less actual winter demand. Any number with a value above 1 means that the market’s predetermined storage fill-to level should be adequate to cover a surge in demand, either in peak winter months, or over the course of the entire winter. As the graph below shows, this winter is very much in line with most winters in the last decade, and in fact above most of them.
As supply moves lock-step with demand, concerns over what is the correct level of storage capacity lessens. What emerges is whether natural gas, whether in storage or via production in the field, can be moved to markets in order to cover new sources of demand, such as incremental power needs or LNG liquefaction. Even at a starting inventory of 3,189 Bcf as of October 31, 2018, the new reality of growing systemic demand is not as much of a threat to storage capacity as previously thought.
Storage capacity is still very important to the market and is destined to become even more important as gas-fired power generation continues to grow. It is deliverability, or more importantly, the ability to inject into storage that will be key. High-deliverability salt cavern storage in the Southeast can be displaced to the Northeast in times of need to serve load. The ability to move gas from place to place, into and out of regions and sub-regions of the Lower 48 that command more gas on any given day, will be of paramount importance in the future.
Midstream operators have been very active over the past several years, building expansion projects to move gas from prolific producing areas to growing demand areas. For example, in calendar year 2018, nine major pipeline expansion projects totaling 7.8 Bcf/d of capacity were placed in-service, with the majority of them moving gas out of the Northeast to destinations in either the Southeast/Gulf Coast or Midcontinent. In 2019, more pipeline expansions will be pursued and constructed to deliver gas out of the Permian in West Texas, in addition to the Northeast, further affecting gas flows and regional storage needs.
As the interplay of demand, supply, new pipeline infrastructure and storage evolves, stay on track by reading OPIS PointLogic's news and analysis.