SHALE gas has transformed the North American natural gas market, but its potential to influence global markets is often called into question. Our analyses predict greater international flows of gas over the next decade, as a result of surging US supply.
Initially, these flows may not affect the existing stratification of international gas markets. But, eventually, volumes can be sufficient to influence and possibly break the world's three-way, regional gas-market model: oil-linked Asia; gas-on-gas North America; and Europe, a hybrid of both. In this scenario, the approximate gas-price ratio of 1:2:4 for North America, Europe and Asia will not hold.
Interestingly, in the 10 years to 2008, natural gas markets were heading towards global convergence at a higher price before North American shale gas took off. The US's natural gas reserves were falling and the country was rapidly building import capacity. Europe's mature gasfields were in decline, and Russian oil-price-indexed gas supplies were increasingly influential, threatening nascent European trading hubs. And surging demand in import-dependent Asia was increasing competition for limited liquefied natural gas (LNG) production, forcing prices up. Although gas markets were not physically connected, planned infrastructure suggested convergence was inevitable - there were long-distance pipeline projects to supply Europe and China; and massive investment in LNG export capacity world-wide. The tightening supply picture prompted the US spot market to jump into line with oil-indexed prices in Europe and Asia.
The unconventional gas boom in North America halted any such convergence. In a few short years, the US became very, very long in gas - long enough to potentially export large quantities - lowering prices to less than one-third of their peak in 2008 (Canadian gas prices witnessed a similar decline).
The question now is whether the impact of North American gas will spread globally. The answer comes down to volume. In the years to 2025, the contribution to supply from international shale gas will be limited. And although rising international LNG supply will contribute significantly to global gas availability, most of these volumes will not have the same cost advantage as gas from North America. So the root question becomes how much North American gas could seek global markets, relative to the size, and residual need, of those markets. If the volume is small, the status quo - three distinctly priced regional gas markets - is unlikely to be disrupted; if it is large, it will be very difficult to maintain.
North American gas supply has become very large relative to local demand. Recoverable resources amount to 100+ years of supply, and new technologies are bringing additional resources into play (see Figure 1). Drilling and production efficiency have steadily increased as the industry has progressed along the learning curve. As a result, break-even costs have fallen by more than 40% in several important gasfields and more plays are becoming economic (see Figure 2). US natural gas production has risen accordingly, reaching a record 24.2 trillion cubic feet (cf) in 2013, up by more than 20% since 2008. Efficiency improvements have enabled such production levels to be met with a fraction of the drilling fleet; in early 2014, the US gas industry was operating with half as many gas-directed rigs as it was in 2012.
It is difficult to see what could stop this gas from coming to market when needed. On the technical side, some subsurface experts have opined that the industry is rapidly exhausting sweet spots and core acreage, and is underestimating decline rates from unconventional wells. Our work indicates that improvements in technology and working practices will continue to expand the economically producible resource envelope, identifying new sweet spots across larger swathes of acreage and blurring the distinction between what the industry currently classifies as core and non-core acreage. In addition, there is now sufficient empirical evidence to rule out a disruptive impact on flow rates and ultimate recovery from decline rates.
Turning to the surface, a prolonged oil price slump would have an impact on associated gas production, but that could quickly be offset by diverting rigs to drill for gas (and drilling rig rates would also be likely to drop, making more gas economic to develop). Regulations theoretically could halt gas drilling, but the states regulate much of their oil and gas activity, and the current carbon-sensitive federal administration is supportive of gas production. Beyond this, the industry continues to have access to land, capital, infrastructure, and service capacity, and remains capable of executing projects efficiently.
On the aggregate, it may seem a bit counterintuitive that declining conventional reservoir production is almost exactly offset by growing shale-gas production, since shale gas wells themselves are known to have steep decline curves and the overall gas rig count has fallen. How is this offset possible and how does this shale-gas production equation work? A combination of factors that have had a positive impact on both the number of gas wells drilled and well productivity has made this possible. As shale-gas production continues to mature, it has been clearly established first of all that more wells are being drilled per rig (ie rig efficiency is increasing with improving techniques such as pad drilling). Secondly, wells are now being drilled in richer and higher-yield spots (new sweet spots are being continually discovered), with higher initial production rates and estimated ultimate recoveries. The dry gas production equation is therefore only getting better over time due to drilling and completion technology and other technical advances. Furthermore, there will be an increase in associated gas - which currently accounts for nearly a fifth of gas production in the contiguous US, as the quest for tight oil steps up. Therefore, North America will likely witness further growth in shale gas production in the foreseeable future.
In the future, at least 1,000 trillion cf of US gas (see Figure 3) and about 500 trillion cf of Canadian gas should be technically recoverable at a price of $5 per million cubic feet (Mcf), even with existing technology, and the industry has the capital and talent to develop it and increase the recoverable total.
The pool of available gas is large, but how does it compare to current and anticipated demand? North American natural gas consumption amounted to about 29 trillion cf in 2013, accounting for over a quarter of total energy demand. Consumption is more or less equally divided between the power, residential and commercial, and industrial sectors. The substitution of coal by gas in the power sector is now an established pattern; in the past few years, about 15% of coal-fired generation has switched over to natural gas. However, there is a limit to how much coal-fired power can be supplanted by gas; lower coal prices and forthcoming environmental regulations will limit the growth of natural gas demand in power to an average of 10-12 trillion cf by 2025. Meanwhile, industrial demand for natural gas has been increasing since its historical dip in 2009. Between then and 2013, industrial-sector demand for gas increased by 20% to 8 trillion cf, mostly as a result of the energy-intensive requirements of iron, steel and bulk chemicals manufacturers. We expect this growth to continue, albeit at a slower pace, increasing by about another 1 trillion cf by 2025. We do not expect natural gas to become a major fuel source for the transportation sector beyond certain niche compressed and liquefied applications because of long technology life-cycles and infrastructure barriers. Gas use already is widespread in North America, so even healthy growth in the industrial and power sectors will not result in a step change in gas demand.
Overall, we expect North American demand for natural gas to be around 35 trillion cf a year by 2025, amounting to a cumulative demand of just above 350 trillion cf over the 2014-2025 period. In comparison, approximately 1,500 trillion cf of gas is available at or below a price of $5 per Mcf, and 850 trillion cf at or below a price of $4 per Mcf. This would imply that more than 1,000 trillion cf of cumulative gas supply through 2025 could be available for export after indigenous demand has been met. Supply could increase yet further as a result of natural gas liquids (NGL) credits, associated gas and technology improvements. Even a small fraction of this volume would be sufficient to flood the LNG market. The supply curve suggests that this kind of volume could be exported without causing a significant increase in local gas prices. There could be temporary strains on the system and spikes in prices due to infrastructure and cold weather, such as those that we have seen in the 2013/14 winter, but these would not much affect the economics of investments in gas consumption, particularly in LNG exports that would bring gas to structurally more lucrative markets.
While the total volume of available North American gas resource is huge, it is equally important to note that the rate at which it can come to market is high, increasing, and flexible. To meet North American demand, local gas production would have to grow at just above 1% annually over the period 2014-2025. If gas production were to grow at the recently observed rate of more than 5%, total production through 2025 would be nearly 50 trillion cf, or 15 trillion cf more than demand. As noted above, most of the drilling fleet is not needed today to meet gas demand. So the rate of extraction for North American gas could be higher today and certainly much higher by 2025. In fact, if we were to return to the same level of gas-directed rigs as 2012, and without accounting for further efficiency gains through technology and workflow which we reasonably expect, North American cumulative production would outstrip demand by more than 100 trillion cf. If sufficient liquefaction infrastructure were available, North America would be capable of exporting gas volumes of this magnitude. Given the potential for such a large excess of supply, it is also difficult to see North American gas prices settling significantly above $5 per Mcf, barring the occasional spike, for a very long time. As a result, cheap feed gas should remain available for LNG export.
Figure 1: Unconventional revolution has left the US with significantly larger gas reserves
Figure 2: Improvement in technology and production efficiency is lowering breakeven prices
Looking to LNG
Global annual demand for LNG, meanwhile, is currently 14 trillion cf (a little over 10% of total gas consumption) and is projected to rise to 22 trillion cf by 2025, for a cumulative production in the 2014-2025 period of greater than 200 trillion cf. Therefore, North American gas could meet over 100% of the expected increase in global LNG demand (see Figure 4). Of course, the US is not the only country that could meet this demand. Qatar already meets a quarter of global LNG demand. Australia is building seven new LNG projects and its LNG production capacity will rival Qatar's by the end of the decade. Mozambique is planning to build LNG trains comparable in size to those in Qatar and Australia. Papua New Guinea will start LNG shipments this year. Other LNG exporters include Indonesia, Malaysia, Algeria, Russia and Yemen. Global LNG capacity would amount to nearly 50 trillion cf a year if all proposed projects were to go ahead - well in excess of projected LNG demand (22 trillion to 34 trillion cf) by 2025. Although not all proposed projects will proceed, capacity is set to rise considerably and exceed demand. Does this leave any room for North American shale LNG? Yes, because of North American gas's cost advantage.
The landed cost in Asia of North American LNG will be the Henry Hub (spot) price plus liquefaction (including conversion margin) and shipping costs. This would amount to roughly $10-$13 per Mcf and would make North America competitive with other suppliers. The actual advantage may be greater, as many recent international LNG projects have incurred substantial capital overruns and delays (see Figure 5). There are 36 proposed export terminals in North America; planned capacity amounts to 12 trillion cf in the US and 8 trillion cf in Canada. Federal regulators have already approved 7.5 trillion cf of this capacity, although only 1 trillion cf has site approval. Nonetheless, the low cost of North American gas production in the field supports a competitive product once chilled and shipped to international markets.
In addition to its structural cost advantage, the nature of gas production in North America provides an additional risk advantage. Gas production is mostly onshore, so less subject to weather disruption. Being onshore also allows North American gas to be brought on in relatively small increments, especially compared to greenfield international projects. Onshore wells in North America are also better adapted to the rapid introduction of technological innovations than offshore wells in other regions. North American gas production is also spread out over a wide geographical area, yet inter-connected by a highly evolved pipeline network, further diminishing the risk of large supply shortages. Collectively, it is easy to see how North American LNG projects have clear risk and cost advantages over most international projects. And Asian customers would welcome supply diversity from secure partners like the US and Canada.
Existing gas-pricing structures look precarious in the face of excess North American gas and planned production from other countries. There are three pricing regimes for major natural gas markets today - (i) hub based spot pricing in North America; (ii) oil-price-linked, long-term contracts in Asia; and (iii) a hybrid of crude-linked long-term and hub-based spot pricing in western Europe. The highest prices are in Asia, which has been perennially short of gas. The Fukushima Dai'ichi nuclear disaster in 2011 made the Asian supply situation worse, as the Japanese authorities shut down the country's nuclear power-generation fleet, forcing the country to switch to gas. Some formerly large LNG exporters, including Malaysia and Indonesia, are becoming net importers. Most of the natural gas supply to Asia is in the form of LNG, and global LNG production capacity has been limited to date. This has resulted in a high degree of uncertainty in relation to the continuity of supply, forcing consumers to enter into long-term contracts at oil-based prices.
We expect global gas demand to grow to 145 trillion cf (or 184 trillion cf in the high case) by 2025, from 120 trillion cf today. A large portion (about 85% in base and about 82% in high case) of the total demand will be met by indigenous production or piped gas. The remainder (22 trillion cf in the base case and 34 trillion cf in the high case) will have to be supplied in the form of LNG, which, as the marginal source of supply, will continue to determine market prices, as it has over the past decade. However, the uncertainty surrounding LNG supply is likely to diminish, in view of the large volumes planned in North America and elsewhere.
Current global LNG demand is 14 trillion cf, but we expect an increase in demand between 8 trillion and 20 trillion cf by 2025. Given the various advantages identified above, North America should supply a large amount of the world's incremental LNG requirements. Our expectation is that North American exports will start in the 2017-18 timeframe, and have the potential to rise rapidly to between 3 trillion and 11 trillion cf a year by 2025. In other words, North America could contribute up to 100% of incremental (base-case) LNG demand (see Figure 7). Even if North America exports are in the middle of the range (around 6 trillion cf), if those plants yet to be built and a debottlenecking of supplies from the rest of the world (5 trillion cf) come to the market, and only half of the proposed non-North American projects are eventually completed, we get more than two times coverage of the base-case incremental global LNG demand and greater than 100% coverage of high-case demand. At this point, supply uncertainty will have transformed into excess supply in search of demand. That is a situation under which markets tend to break.
Figure 3: US natural gas supply
Figure 4: Global LNG demand and North American gas resource potential
Figure 5: North American LNG landed cost to Japan (assuming 15% IRR liquefaction)
Figure 6: North American LNG will enjoy a favourable position in the LNG supply curve
A shift in supply
Change of this kind will not happen overnight. North America will initiate the shift from supply uncertainty to excess, but there will be other contributors to supply, such as countries in Africa with abundant stranded gas resources. In other words, the gross imbalances will not be resolved neatly, but by negotiations between many companies and many governments, requiring massive expenditure on gas production, liquefaction and transportation. Many investments will likely not move forward for financial or regulatory reasons and some of those that do progress will experience cost and timing overruns. Even North America will face difficulties when its exports exceed 20% (or about 6 trillion cf) of local consumption: the temporary strain of balancing local consumption with a high level of exports may cause short-term prices spikes and regulators may become trigger happy and limit exports. However, despite all these uncertainties, there will be enough gas in the pot to stir and instigate a new era of convergence.
The new order will evolve in phases. In the first phase, sufficient LNG volumes will be made available to the market to begin to break the link with crude oil prices in supply contracts. We are already seeing the initiation of this phase in cases where volumes are contracted from the US. As volumes increase, spot trade will become more liquid. This will trigger the second phase, in which price volatility will diminish, encouraging more spot buyers to participate in the market. As trade expands, conversion and transportation costs will benefit from economies of scale, and spreads will be narrowed. This will usher in the final phase: hub-based gas-on-gas pricing (see Figure 8).
This shift will be driven by sizeable financial incentives. In terms of absolute price differentials, we believe the incentive could eventually be significant. By the time the final phase is complete, which may be after 2025, gas-price ratios could be 1:1.75:2.5 (North America:Europe:Asia). Given our expectation for Henry Hub prices, this implies that the spread between Asia and North American prices could shrink by 50% (or more) from the current level. Utilities in Japan and Korea will start seeing this phenomenon in the coming decade and will be unable to resist buying in the spot market. They may even begin to revise existing contracts, and encourage the development of trading hubs to replace long-term, take-or-pay contracts. This happened during the deregulation of US gas markets: before the mid-1980s, the US gas market was highly regulated - not unlike the global gas market today. Following Federal Energy Regulatory Commission's deregulation of the gas market, large amounts of gas became available, catalysing the development of a spot market in which prices were significantly below those in the long-term, take-or-pay contracts that utilities then had with gas suppliers. The utilities rejected their contracts en masse, claiming a form of force majeure. It was a multi-billion dollar debacle, but one that resulted in the cancellation of take-or-pay contracts and a universal switch to the spot market. The same incentives will arise in Asia over the next decade and some of the same players will face the same choices.
Even if a cartel of potential suppliers of incremental LNG to Asia were to be formed, it probably would not prevent the break-up of the current market stratification and would likely be illegal. There is enough uncertainty in the ability of others to execute, in demand forecasts, and national interest, to move ahead with a large part of the planned investments. Newly discovered North American gas would become available not through a single entity, but through hundreds of producers, pipeline operators, plant operators and ship owners. Such an intricate cartel of producers, chillers, and shippers is unlikely to be formed.
The imbalance that we see developing is mostly to be expected. Natural gas is a necessity to modern economies. That we will find and develop the exact amount needed, not more and not less, is unlikely. The availability of unconventional gas in North America has tipped the balance domestically and may have a similar effect on global markets - to say nothing of the prospects for unconventional gas outside North America. Now that the innovation in gas recovery has become clear, the imbalance and resulting market stress becomes more a matter of timing. And timing is critically important in a capital-intensive industry like oil and gas. Consequently, public and private decision-makers will need to gauge the implications of a market shift. For example, if large volumes of LNG from North America can make their way to the global market, how should they adjust their plans for unconventional and frontier development globally? Should they, in fact, amend these plans at all? What about LNG development in regions where the cost of developing LNG projects is high, such as Australia and East Africa? Will chemical and energy-intensive manufacturing companies slow their move into the US even if it stays cost advantaged but not by much, and maybe even cost neutral after accounting for moving finished products to the demand centres in Asia? What will the various countries with ambitious plans to develop renewable resources do? Will cheap gas deal clean-tech another blow? Will the growth of coal-fired power in places like China stall? Questions such as these form a complex strategic picture for players across the energy value chain.
Ultimately, if the resource exists - and we know it does, in abundance - the industry will find a way to get it to the desired consumption points. Consequently, convergence of the gas markets, excluding transportation basis and other structural factors, is not a question of "if' but of 'when' - and based on the evidence, it may very well begin to happen in the next decade.
Figure 7: North America well positioned to cover large portion of incremental LNG demand
Figure 8: Current oil-linked long-term contracts will eventually break