To illustrate this principle, we compare a simplified model of a conventional oil field with a comparable model tight oil field. We model the conventional oil field development as a series of 48 wells, completed at the rate of one per month. Each well has an initial (maximum) production of 1000 bbl/d, performance which is above average but not unknown in U.S. onshore fields. Following standard oilfield practice (Cosse, 1993), the field is assumed to be put on secondary recovery immediately after production starts, thereby maintaining reservoir pressure.
We model tight oil field development using assumptions similar to those used for the model conventional field: 48 wells, completed at the rate of one per month, with initial production of 1000 bbl/d, again above average but not exceptional (Sandrea, 2012, EIA, 2016a). Because tight oil fields cannot normally be put on secondary recovery (Kleinberg, 2014), individual wells decline rapidly in the first several years, typical of primary recovery.
For an ensemble of wells completed at times tk, with k ranging from 1 to N, where N is the total number of wells completed, the rate of oil production from the field at any time t is given by
where qk(t-tk) is rate of production from a single well k at the time t subsequent to the completion of that well at time tk. Since wells do not produce prior to being completed, qk= 0 for all t < tk. Eq. (1) allows each well to have a unique decline curve qk. In our models we assume all conventional wells have a common decline curve, qc, and all tight oil wells have a different common decline curve, qt.
In a conventional oil field under secondary recovery, rates of decline are roughly uniform over much of the life of each well:
where αy is the annual rate of decline, which we shall assume to be αy = 0.06/yr. This corresponds to an annual rate of decline of 6%, a value that is justified below. The monthly rate of decline is αm = αy/12. This simple differential equation is integrated to find the conventional oil well decline curve when the field is on secondary recovery; IP is the initial production rate of an individual well:
For the model of a tight oil field, we assume that all wells have a common decline curve qt, given by a Bakken average type curve (IHS, 2013b) normalized to an initial production rate of 1000 bbl/d, see Fig. 3.
After the cessation of completions in month 48, the conventional oil field declines at an annual rate of 6%; a sum of exponentials decays at the same rate as the individual exponential functions of the argument of the summation. With this knowledge, we selected the individual well decline rate, αy = 0.06/yr, based on a global average of conventional oil field decline rates (IEA, 2013).
The results of the two models are shown in Fig. 6. During months 1 to 48, while wells are being completed, the production from both fields increases with time. Because the conventional wells decline rather slowly, the ramp up of production during the development phase is nearly linear. The much more rapid initial decline of production of the tight oil wells leads to a distinctly sublinear ramp up of production. This is the origin of the “Red Queen Race” (Likvern, 2012).
Fig. 6. Modeled field-level production in conventional and tight oil (Bakken) fields, during and after 48 month drilling and completion campaigns, using the individual well decline curves shown in Fig. 3.
After the cessation of completions in month 48, the conventional oil field declines at an annual rate of 6%, the global average of conventional oil field decline rates. Unlike the conventional oil field, the tight oil field does not decline at a time-invariant rate following the cessation of drilling, as shown in Fig. 6. Table 3 provides a summary of annual production decline rates of the model conventional and tight oil wells and fields. Although tight oil fields experience a substantial decline in production in the first two years after cessation of drilling, as the most recently-drilled wells decline, a larger number of slowly-declining legacy wells supports substantial continued production.
Thus tight oil fields with large legacy inventories of wells will produce substantial quantities of oil for many years after completions have ceased. Note that Table 3 is only illustrative: tight oil field decline rates depend on details of the development schedule. If completion activity has increased immediately prior to cessation, a large proportion of wells in the field are relatively new, leading to faster initial decline of field-level production once drilling and completion comes to an end. On the other hand, if completion activity has slowed in the year or two before terminating, production after termination will decline more slowly than suggested by Fig. 6 and Table 3.
Table 3. Percentage annual decline of conventional oil well and field under secondary recovery, and tight oil well and field under primary recovery. The field level declines follow the termination of the drilling and completion program. The tight oil results are model-dependent, as explained in the text.
It is generally agreed among oilfield service companies that hydraulic fracturing is an imperfect method for connecting gas or oil in low permeability formations to the wellbore; according to production logs, 30–40% of fractures do not produce fluids (Jacobs, 2015, Hunter et al., 2015). Either of two types of refracturing are used to overcome this problem. In the “reconnect” procedure, fractures that are poor conduits for fluid flow can be reopened, whereas in the “restimulate” procedure, new fractures are created (Hunter et al., 2015).
Refractures cost between about 20% and 40% of the cost of a new well (Lindsay et al., 2016), so it would seem the economic case for these techniques would be compelling. Nonetheless, experience has shown that a rigorous screening process maximizes the chance of success (Hunter et al., 2015), and only a few percent of shale gas and tight oil wells drilled in the last ten years have been refractured using chemical diversion, the most cost-effective technique (Lindsay et al., 2016). Lack of predictability is the main barrier to widespread implementation (Wood Mackenzie, 2016a).
Infrastructure, labor and financial inelasticities
Following an industry collapse, as occurred in 2014–2016, the rate at which tight oil production can ramp up once drilling resumes depends on how equipment was taken out of service during the period of low activity. If equipment is written off, it is destroyed or cannibalized. However, when equipment is stacked, as is the practice of some large service providers (Schlumberger, 2015, Seeking Alpha, 2016), it is assumed to retain value as a productive asset and is warehoused accordingly.
A second factor is labor availability. Labor required in the tight oil sector, along with associated equipment, made a smooth transition from gas drilling to oil drilling in 2009. Following massive layoffs from the petroleum industry in 2015 and 2016, skilled labor may not be as abundant in the future as it has been in recent years. The duration of training varies with the degree of skill and specialization required, and can exceed a year to gain proficiency in some job categories.
Financial markets also introduce inelasticity. The ready availability of capital played an important role in the initial growth of the US tight oil industry, with many producers, year after year, operating at negative free cash flow (cash flow after capital investments) (Sandrea, 2014, EIA, 2014, EIA, 2015d, Domanski et al., 2015). It remains to be seen whether debt and equity financing is as available in the future.
Oil market stability
Short term market stability
Although it has been stated that US tight oil can challenge Saudi Arabia as the world’s marginal producer (e.g. The Economist, 2014), this assertion is open to question. Spare capacity is the most important characteristic of a swing producer. Spare capacity is defined as production that can be brought on line within 30 days and sustained for at least 90 days (EIA, 2016f, Munro, 2014). While there is no doubt Saudi Aramco can increase production this rapidly, the US tight oil industry cannot. In addition, unlike OPEC members, who can in theory increase or reduce their oil production in concert, the hundreds of U.S. producers cannot and will not coordinate their activities.
Inventories of crude oil and petroleum products are also drivers of short term market stability. As of the first quarter of 2017, US commercial crude oil and product inventories amounted to 1.34 billion barrels, with an additional 0.69 billion barrels of crude oil in the U.S. Strategic Petroleum Reserve (SPR) (EIA, 2017a). Altogether this amounts to about 100 days of U.S. consumption. If needed, the SPR can be drawn down at a maximum rate of 4.4 million barrels per day for 90 days, after which the maximum rate decreases (Carr, 2017). Thus SPR meets the definition of spare capacity.
Medium term market stability
Although tight oil resources do not constitute spare capacity in the strict sense of that term, it is useful to consider a second type of spare capacity: a medium-term spare capacity, which can be brought on line in a few months. Given the special circumstances explained in Section 4.1, the U.S. tight oil industry sustained production rate increases of 1 million barrels per day per year from January 2011 until January 2015, see Fig. 1. Other factors also suggest that U.S. tight oil production can contribute to medium term price stability.
Drilled but uncompleted wells
Drilled but uncompleted wells (“DUCs”) have become a resource capable of providing medium-term market stability. The use of the “drilled-but-uncompleted” terminology is widespread but imprecise; such wells should be called “drilled and partially completed”. Completion of oil and gas wells includes lining the well with steel pipe, cementing the pipe in place, perforating the pipe to either start hydrocarbon production or to create fracture initiation points, stimulating the well (most often by fracturing), and installing other equipment (such as pumps) to bring liquids to the surface.
A DUC has been drilled, cased, and cemented, but perforation and stimulation have not yet been performed. The cased and cemented well is stable, and production can be delayed indefinitely without risk. Once the well is perforated and stimulated, reservoir equilibrium is disturbed and stopping the flow of fluids can have unintended consequences that can negatively affect future production.
There are several reasons why more wells are drilled than are completed (Nasta, 2016). First, delays in scheduling and mobilizing fleets of high-value hydraulic fracture equipment are normal. Second, some drilling rigs are leased on long term contracts, which are uneconomical to cancel prior to the completion of a drilling campaign. Third, some leases require a certain level of drilling activity, or the lease is forfeit. Fourth, a lag in the availability of pipeline capacity affects the rate at which gas wells, or oil wells that produce substantial amounts of associated gas, can be put on line. Fifth, after the fall in the price of oil started in mid-2014, fracturing was delayed in the expectation that oil left in the ground would be worth more in the future.
Some of the reasons for delaying well completion are illustrated by Fig. 7. When prices decline (July 2014–January 2015; September 2015–January 2016), the inventory of DUCs grows, as operators keep oil in the ground in expectation of higher prices in the future. DUC inventory also grows when activity is strong (January 2014–July 2014; January 2017–April 2017) due to infrastructure and service bottlenecks.
Fig. 7. U.S. tight oil and shale gas drilled but uncompleted wells (EIA, 2017b) (dotted curve, right axis) and West Texas Intermediate crude oil price (EIA, 2017c) (solid curve, left axis) from January 2014 to April 2017.
Leases held by production
Markets can also be stabilized by the availability of undrilled well locations, where the resource has been de-risked and leases have been secured. The desire of oil and gas producers to have a substantial inventory of locations ready to be drilled in response to market signals can conflict with the land- or resource-owner’s desire to realize royalty income as quickly as possible. These interests are balanced by lease contracts, which typically provide that a lease granted to an oil or gas driller is cancelled unless royalties are derived from it within a specified period, often three years. The lease remains in force for as long as the stream of royalty revenue continues (Smith, 2014, Herrnstadt et al., 2017).
Leases must encompass the entire subsurface volume drained by wells drilled on them. Since shale gas and tight oil wells have horizontal legs ranging from one to two miles in length, with perpendicular fractures hundreds of feet in length, leases are now commonly one to two square miles in area. Each lease may include property from several resource owners (“unitization”), with royalties pooled and divided equitably.
In order to maintain maximum operational flexibility, an oil or gas producer may drill, complete, and put on production only a single well in a lease that might be fully developed by six or more wells. That well might remain on production for many years, albeit at low levels. Such leases are held by production, and allow a producer to retain a large inventory of ready-to-be-drilled well locations (Smith, 2014, Herrnstadt et al., 2017).
U.S. tight oil as a price maker
Given the factors promoting medium term price stability, despite important inelasticities, and despite lacking any coordination among suppliers, U.S. tight oil has the potential to impose some discipline on crude oil pricing. The threat of significant new quantities of product entering the market when the price of oil exceeds the lower bound of tight oil full cycle breakeven, about $50–$60/bbl in 2017, may provide a restraint on the expectations of market participants who seek to raise prices by cutting production.
The effectiveness of U.S. tight oil to be a price stabilizer was tested in early 2017. In November 2016, in an effort to increase the price of crude oil, ten OPEC and eleven non-OPEC nations agreed to cut oil production by 1.81 million barrels per day. By April 2017, this group had reduced production by 1.73 million barrels per day relative to October 2016, substantially achieving their objective. However, this effort was partially undercut by the U.S. tight oil industry, which between December 2016 and May 2017 increased production capacity by more than 300,000 barrels per day (EIA, 2017b).
The ability of the U.S. tight oil industry to be a price maker is in marked contrast to industry segments which exploit deepwater, arctic, and other challenging resources. Adding significant capacity in those sectors can take years, and they are thus price takers. The same is true of tight oil resources outside the U.S., as those resources are undeveloped and therefore lack the ability to increase production at a rate that would be significant in world oil markets.
The longer term outlook is less certain. From 2011 to 2015, 1 million barrels a day of production was added each year, a rate of increase almost unprecedented in the history of the industry. However, the U.S. Energy Information Administration Annual Energy Outlook 2016 (AEO2016) reference case predicts a much slower growth rate of about 120,000 b/d per year between 2020 and 2030. Production is expected to increase faster in the High Oil Price case, but then cannot be sustained after 2025. Only in the High Oil and Gas Resource and Technology case do the rate increases of 2011–2015 continue past 2020 (EIA, 2016j).
Given knowledge of a range of breakeven points for a relatively high-cost resource, a lower-cost competitor with ample spare capacity might be tempted to increase production to the extent that the price of the resource falls below the breakeven point range of its higher-cost rival. To be successful, this strategy requires an understanding of the tiered nature of breakeven points. Frequently, breakeven point data are presented by analysts, or in corporate presentations to investors, without adequate disclosure of what exactly is meant by breakeven. In this paper we have shown that benchmark and breakeven points are only useful to the extent their calculation is transparent.
In projections of the reaction of oil production to changes in the price of oil, many analysts underestimated the dynamic nature of tight oil economics. In mid-2014, full cycle breakeven points for U.S. tight oil produced by horizontal well construction and massive hydraulic fracturing were generally in the range of $60–$90/bbl, giving rise to expectations that the price of oil was unlikely to fall below about $60/bbl. Half cycle breakeven points were in the range of $50–$70/bbl, and lifting costs were below $15/bbl.
When oil prices declined, not only did these brackets move to lower cost ranges due to internal and external drivers discussed in this paper, but there was a large-scale transition of breakeven points. Oil production promptly shifted from full cycle projects to the half cycle economics of drilling to maintain level production. After the second half of 2015, drilling was no longer adequate to maintain level production, but oil production continued from pre-existing wells, the rate of production from which falls rather slowly after the first two years or so.
In this paper we propose a consistent treatment of breakeven points – full cycle, half cycle, and lifting cost – and explain internally-driven and externally-driven changes in breakeven point economics. In a rapidly evolving industry such as tight oil production this analysis is itself subject to change. Nonetheless, it is important for a variety of reasons:
- The various levies imposed by governments, including leases and royalties, are calculated to maximize payments while allowing oil producers to retain sufficient profit to make resource development attractive. A better understanding of breakeven points by governments would facilitate this process.
- Asset valuation depends critically on estimates of future costs. Because stable tight oil production requires the continuous drilling and completion of wells, the economics of a long-lived play requires understanding how half cycle breakeven points change over time. Similarly, the economics of growing oil volumes requires analysis of full cycle breakeven points. In both cases, secular changes due to internal and external drivers should be taken into account.
- Energy analysts, in both private and government sectors, can improve forecasts by incorporating into their economic models realistic ranges of breakeven points, and models of how these change under various conditions.
Inelasticity in the response of oil production to market signals is a further complication, the understanding of which requires close examination of individual well decline curves and their implications at play level. The interaction between productive capacity and technical features of oilfield practice, such as drilled but uncompleted wells and refractured wells, must also be considered, as does oilfield and takeaway infrastructure, capital markets, and labor factors.
Breakeven points are among the most useful measures of the economic viability of a hydrocarbon development project. They are particularly useful in assessing the robustness of the project with respect to a decline in the price of the produced commodity: if future market prices are projected to be comfortably above the breakeven cost of a project, the investment is likely to be a profitable one.
This work explores the following characteristics of breakeven point analysis:
- The breakeven point is most useful when its calculation is transparent. We argue that the purveyors of breakeven point data have a responsibility to carefully define the elements that go into their calculations.
- While recognizing that various users will want to define breakeven points in ways most useful to them, we propose a model breakeven point scheme that incorporates many elements of these calculations, and which approximates consensus schemes used by analysts and industry participants.
- We define and explain each major category of expense in our slate of breakeven costs.
- We divide our slate of costs into three major categories: full cycle, half cycle, and lifting cost. These terms are common in the breakeven analyses found in the work of analysts, agencies, and oil producers; we provide precise definitions of them.
- We show how breakeven costs change due to internally-driven factors. These are microeconomic factors which represent normal improvements in operational efficiency. Efficiency improves over time in every new play exploited by the petroleum industry, as geological knowledge and infrastructure maturity progress. Because tight oil production technology has developed rapidly in this decade, internally-driven improvements have been very pronounced.
- Externally-driven factors, driven by the macroeconomic environment – principally the price of oil – also have a strong effect on breakeven point dynamics because, counterintuitively, changes in the market price of oil lead the cost of producing oil.
- Not only do individual costs change due to macroeconomic factors, but the breakeven point structure itself changes. As prices fall, full cycle economics gives way to half cycle economics, and eventually to lifting cost economics. In rising markets, the reverse is true.
We have discussed other factors affecting oil market stability:
- The surprising speed with which U.S. tight oil production increased after 2010 is explained, in part, by the timely availability of specialized oilfield equipment built between 2004 and 2009 to exploit shale gas.
- The unexpectedly slow decline of U.S. tight oil supplies during the oil price declines of 2014–2015 is shown to be due to the relatively slow decline of production from legacy tight oil wells, when considered on a field-average basis.
- Infrastructure, labor, and financial inelasticities also affect oil market dynamics.
- U.S. tight oil production does not ramp up or down quickly enough to significantly affect short term oil market stability. However, it does have the potential to stabilize oil markets in the medium term. Thus unlike deepwater, heavy, or arctic oil resources, which are price takers, U.S. tight oil is likely to function as a price maker in the medium term.
The authors wish to thank Vello Kuuskraa, Kenneth Austin, Rusty Braziel, Imelda Foley, Joel Enderlin, Jan Mares, and the anonymous referees for their helpful comments and suggestions. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.