Read Our Latest Energy Market Insights – Go There >>

2Q 2019 – Spotlights

  • Evaluating the Risks in Inventory Estimates

    BTU Analytics calculates remaining inventory by breakeven across all major unconventional basins in order to understand the marginal economics of the future supply stack. Our inventory model relies on sub-location average well spacing and lateral length assumptions, which dampen the effects of isolated, operator-specific variances.

    Inventory estimates from E&P company investor relation materials are predicated on specific spacing and lateral length assumptions often found in the footnotes, but not all assumptions are created equally. Reeling from the industry-wide newfound vigilance regarding parent-child interference, several producers have announced wider spacing between horizontal wells attempting to improve the productivity of infill wells. In response, BTU Analytics highlighted the effect of the spacing assumptions used in our inventory model for the Permian Basin in the previous edition of the E&P Positioning Report. Taking that analysis a step further, this quarters’ report seeks to enumerate the risks to inventory estimation across all basins in the E&P Positioning Report.

    When E&P companies disclose inventory estimates, common disclosures are the count of undrilled locations and the number of years it would take to deplete this inventory. These metrics often rely on static assumptions for the spacing between laterals, lateral length, and the rate at which new wells will be drilled. Recent results pointing to parent-child interference have challenged the validity of the current spacing assumptions for a number of operators, challenging the reliability of their inventory estimates. Lateral lengths are growing in most basins as drilling efficiencies increase, enabling greater reservoir contact with a smaller surface footprint. The rate of drilling new wells is dictated by capital allocation, which is heavily dependent on commodity pricing. In short, inventory estimates are only as credible as the assumptions they are based on, and there is a considerable amount of flux in these parameters today.

    Quantifying Inventory Assumption Differences

    Management of EQT, the largest gross gas producer in the US, has expressed the desire to change the company’s well spacing in the coming years. In the months preceding the transition of management that took place earlier this month, the public dialogue between EQT Corporation and the Rice brothers illustrated the growing uncertainty stemming from inventory assumptions. Spacing for this year will be 880 feet, but management has indicated that optimal spacing is 1,000 feet.


    The degree of confidence in inventory assumptions is intertwined with well performance. In order to maximize reservoir drainage, the presence of parent-child communication would be a positive result, indicating minor overlap of the drainage areas; however, well-level economics begin to suffer as interference increases due to lower productivity from the infill (child) wells. Following years of downspacing, E&P companies are starting to have sufficient empirical data to evaluate the productivity of wells from their spacing tests. The imperative question is this – after years of experimentation with downspacing, are past well results indicative of future well performance for nearby wells?

    Permian Well Productivity

    BTU Analytics has increased our inventory model’s granularity with the recent inclusion of horizon-level breakeven assignments. Until this report, BTU assigned breakeven values to undrilled locations in the Permian based on the average breakeven of all wells contained in the same 3×3 mile grid box, irrespective of formation. The methodology was useful before widespread delineation of multiple formations across the basin had occurred. But with the growing number of data points covering the basin, a more granular methodology is warranted.  Now, the breakeven assignment is the average of wells that share the same horizon within the grid box. By matching average breakeven values to undrilled locations for each horizon, the accuracy has increased, but the criteria for matching economics to inventory locations is stricter, leading to the increase in the number of locations in the “No Economics” sample size.
    BTU Analytics analyzed the well productivity as measured by initial production rates of crude oil of 200 wells from the most actively drilled 3×3 mile grid boxes from the Delaware and Midland Basins.  Two grid boxes were selected from each of the five most frequently drilled counties in each basin based on 2018 activity. Permian breakevens are most sensitive to the revenue from crude sales, especially in recent months with weak Waha natural gas prices, so oil IP rates serve as a proxy for well productivity. The productivity of the first five wells drilled in each grid box was compared to the most recent five wells drilled in that same 3×3 mile grid box and formation.

    The productivity of recently drilled Midland wells exceeded that of the first wells drilled in each grid box in most cases. The average oil productivity improvement for the Midland Basin was nearly 33% across the 90 square miles represented by the sampled grids. The Delaware wells did not mirror similar productivity improvement. The average change in Delaware well productivity over time was just shy of a 4% decrease, indicating that recently drilled wells failed to exceed the productivity of prior wells targeting that formation in the same grid box. Of the 1,907 grid boxes in the Permian, this analysis focused on the most active 20, ten from Midland and Delaware. Despite the small sample size, the Delaware productivity results were consistent when twice as much acreage was incorporated in a subsequent Delaware well analysis.

    Breakeven economics have not necessarily suffered proportionately to the productivity changes discussed above because well costs per lateral foot decreased significantly in the 2015 – 2017 period. With equity markets pressing for greater capital discipline, a renewed importance to make every well count in all basins has driven spacing revisions for many operators, especially those observing lower productivity from their infill wells. As with all experiments, the delineation of shale acreage is a process, not an exact science.


    Inventory Risks by Region

    Fundamentally, all basins face diverse risks to inventory estimates. Oil-focused regions enjoy the luxury of stronger commodity pricing in today’s market, but economics outside of core areas hasn’t been tested in recent years. In the Permian, the scale of the resource is so great that small changes to assumptions can have major implications for both oil and gas markets. Limited capital investment in the PRB has constrained the horizon delineation process, yet producers like Devon and Chesapeake continue to praise its potential. Bakken laterals are consistently drilled to at least two miles and the core is heavily developed, which can challenge the model that BTU Analytics uses to determine whether enough undrilled land is present for a future location. Oklahoma’s SCOOP and STACK inventory uncertainty stems from the disputed profitability of wells given the limitations of publicly available data.

    Commodity pricing also plays a role in the certainty of inventory in natural gas-focused regions as prices remain near multi-year lows. Activity in the Fayetteville has stagnated due to challenged economics, casting doubt around the true economic feasibility of the remaining inventory. Severe stakeholder opposition to new midstream infrastructure projects in parts of Appalachia and along routes that could transport Appalachian gas are likely to limit future growth from the region. Long laterals are also being drilled with increasing frequency in the Northeast, echoing the challenge of well placement found in the Bakken.

    The primary risk to inventory estimation in the Haynesville is the prevalence of privately-held E&P companies, which often refrain from publicly sharing spacing or drilling program data. However, after the completion of Comstock’s acquisition of privately-held Covey Park, more data may make its way to the public sphere. While less uncertainty stems from delineating new formations in gas plays compared to their oilier regions, lack of activity or data in some plays risks obscuring inventory estimates.

    Inventory estimates published by E&P companies rely on a mishmash ofassumptions, making aggregation difficult. As a counterpoint, BTU Analytics’ inventory model employs more generic spacing and lateral length assumptions, which insulates the resulting inventory calculations from operator-specific spacing deviations. The assumptions that underlie our location estimates are also re-evaluated for each edition of the E&P Positioning Report, to ensure that recent trends are accounted for.


    Lack of Firm Transport to the Gulf Presents Haynesville Operators with Basis Risk

    As oil and gas production has grown across the US, infrastructure can fail to keep pace and creates bottlenecks that producers may be unable to escape. When these bottlenecks surface, as they have in the Haynesville, in-basin pricing can weaken and producers must decide whether to rein in production growth ambitions. One factor that can differentiate producer strategy is taking firm capacity on pipelines. With firm transport, operators may be able to sell their gas into markets not affected by bottlenecks, thus making infrastructure a differentiating factor for operators active in a basin constrained by pipeline bottlenecks.

    Haynesville gross gas production has increased rapidly over the last two years, nearly doubling since late-2016 as the chart below shows. At the same time, this production growth has been met with increased inbound gas flows from other growing regions, namely Appalachia and Oklahoma. Appalachian gas has been increasingly flowing south towards the Gulf Coast in non-winter months, as flows into the Midwest on Rover and Nexus need to find a home as demand in the Midwest lulls once the winter ends. Additionally, growing gas production from Oklahoma has also been flowing towards East Texas and Louisiana markets on the Midcontinent Express (MEP) and Gulf Crossing pipelines.

    This production growth makes sense, in theory, as growing demand for gas from new LNG facilities has made the Gulf Coast a viable market to send new gas molecules. However, a lack sufficient pipeline capacity from Carthage (East Texas) and Perryville (Northeastern Louisiana) to the Gulf Coast has limited the ability of gas to move efficiently to the Coast. The result is a weakening in local pricing, as shown in the chart above, that BTU Analytics forecasts to remain weak into 1H 2022.

    With in-basin pricing weak and expected to remain so for the next several years, Haynesville drilling activity could slow until pricing improves. The chart on the next page shows Haynesville drilling activity from 2015 to present. Aethon, Indigo, Rockcliff, GeoSouthern and Vine make up the top private operators in the Haynesville while Comstock, ExxonMobil, BP and Chesapeake are the most active public operators in the region. Together, these operators represent 76% of Haynesville drilling activity in 2019. While these operators may currently be the most active operators, pipeline data indicates that many of these operators don’t hold significant firm transport capacity on pipelines and thus might be exposed to basis risk.

    The chart below shows the latest available gross gas production (October 2018) by operator in East Texas and Northern Louisiana as compared to firm transport commitments for those operators on interstate pipelines that would be able to transport gas through the existing bottleneck between local markets and the Gulf Coast. This list excludes intrastate pipelines, like Acadian, which do not have the regulatory requirement to post an index of customers, and therefore isn’t a complete accounting. With the exception of BP, Range Resources and Shell, which appear to have firm transport to market in excess of their gross production, most of the most active producers do not have firm transport commitments in the public data to support their current gross gas volumes. Notably, operators could have entered marketing or firm sales arrangements with companies that have firm transportation, which could lessen basis risk exposure.  These arrangements are rarely made public.  However, several companies have highlighted their lack of transport obligations, including Comstock at the time of the Covey Park acquisition announcement and Vine in a 2017 filing.

    While Acadian does not disclose its index of customers, gas flows indicate that Covey Park (now Comstock), BP, and EXCO volumes are on that pipeline. However, whether those volumes are moving under firm transport arrangements is unknown.

    The firm transport data is just one piece of this puzzle, but it indicates that a number of Haynesville operators are likely exposed if in-basin pricing moves weaker. BTU Analytics forecasts that completion activity will begin to slow in October 2019 based on this analysis. To the extent that Haynesville and Cotton Valley producers do not respond, though, Carthage and Perryville pricing could weaken beyond current estimates, further pressuring operators without firm transport. See below for a detailed list of the firm transport commitments discussed in this analysis.

Share This Article

Share on facebook
Share on twitter
Share on linkedin

Recommended for You

Log In

Energy Market Insights

Receive Free Energy Market Insights When They Are Published