Permian Basin Decline Curve Benchmarks: What Public Data Reveals About Type Curve Accuracy in 2026

Editorial disclosure: This article reflects the independent analysis and professional opinion of the author, informed by publicly available production data, regulatory filings, and peer-reviewed petroleum engineering research. No operator, service company, or data vendor reviewed or influenced this content prior to publication.

The Permian Basin produces over 6.5 million barrels of oil per day. Every one of those barrels sits behind a production forecast, and those forecasts drive decisions that collectively move billions of dollars a year: reserves bookings that determine borrowing base capacity, acquisition valuations that set deal prices, and drilling programs that commit hundreds of millions in capital. When the type curve is wrong, everything downstream from it is wrong.

And the type curves are wrong more often than the industry admits.

This article examines what publicly available production data actually reveals about decline behavior in the Permian's major formations, how operator type curves compare to realized performance, and what that divergence means for engineers, executives, and investors evaluating Permian assets in 2026. The goal is not to replace your reservoir engineering team's analysis but to give you a framework for stress-testing it.

Why Type Curve Accuracy Matters More Than Ever

Type curves are the atomic unit of E&P valuation. A single type curve feeds into three categories of decisions that cannot tolerate systematic error:

Reserves booking. SEC-compliant proved reserves require "reasonable certainty" that quantities will be recovered. Proved undeveloped (PUD) reserves are booked against type curves for locations that have not yet been drilled. If your type curve overstates EUR by 15%, your PUD reserves are overstated by a comparable amount, and your borrowing base is inflated accordingly. The reserves auditor catches this eventually, but the damage to credibility and capital access happens fast.

Acquisition economics. In the 2025-2026 A&D environment, where U.S. upstream M&A reached $65 billion in 2025 and operators like Permian Resources have signaled up to $3 billion in Delaware Basin acquisition capacity, the type curve is the single most consequential assumption in a deal model. A 10% error in the type curve EUR on a 500-location inventory translates to hundreds of millions in misvalued reserves. ConocoPhillips evaluating a potential $2 billion Permian divestiture, SM Energy targeting $1 billion in post-Civitas merger asset sales -- every one of these transactions hinges on the buyer's and seller's type curves being grounded in reality.

Drilling capital allocation. When a development team ranks 200 locations by rate of return to build a three-year program, the ranking is only as good as the type curve behind each location. If Wolfcamp A type curves systematically outperform Bone Spring type curves but development costs differ, the capital allocation sequence changes. If Tier 1 and Tier 2 locations use the same type curve, the program's aggregate economics are overstated.

In a basin where Tier 1 inventory is 60-65% depleted and operators are increasingly drilling into less productive rock, the margin for type curve error has narrowed to the point where it directly affects corporate strategy.

Public Data Sources for Decline Analysis

Before diving into formation-level benchmarks, it is worth cataloging what data is publicly available and what its limitations are.

Texas Railroad Commission (RRC). Monthly production by lease and well, completion data, and well logs. The RRC is the primary source for Texas Permian wells. The main limitation is reporting lag: production data is typically 2-4 months behind, and early months are frequently revised upward as late-reported volumes are allocated. Any analysis using RRC data must apply a production adjustment factor to recent months or exclude them entirely.

New Mexico Oil Conservation Division (OCD). Monthly well-level production for the New Mexico side of the Delaware Basin. Data quality has improved significantly over the past several years, though formatting inconsistencies still require cleanup.

Enverus (DrillingInfo). The dominant commercial database, aggregating state-level data with proprietary normalization, lateral length estimates, completion parameters, and pre-built type curve tools. Enverus data is the de facto standard for A&D analysis. The limitation is cost and the black-box nature of their normalization algorithms. If you are benchmarking type curves, you need to understand how Enverus adjusts raw state data.

Novi Labs. Provides public Permian Basin analytics including productivity-per-foot trends, spacing analysis, and operator benchmarking. Their blog publishes semi-regular updates on basin-level performance trends with enough granularity to validate or challenge internal type curves.

EIA Drilling Productivity Report. Monthly estimates of new-well production per rig and legacy production decline by basin. Useful for macro-level context but too aggregated for formation-level type curve work.

Operator investor presentations. Quarterly earnings decks from public operators (Diamondback, Permian Resources, Devon, Pioneer/ExxonMobil, ConocoPhillips) typically include type curves by formation, well cost assumptions, and occasionally actual-vs-type-curve comparisons. These are invaluable for benchmarking, but remember that operators have an incentive to present type curves that support their capital program narrative.

The key principle: no single source is sufficient. Cross-referencing state regulatory data with commercial databases and operator disclosures is the only way to build a reliable picture.

Permian Basin Decline Behavior by Formation

Wolfcamp A: Delaware Basin

The Wolfcamp A in the Delaware Basin has been the primary development target for operators including Permian Resources, ConocoPhillips, and Occidental. Wells here are characterized by high initial productivity -- IP30 rates in core areas of Loving, Reeves, and Ward counties have historically ranged from 1,200 to 2,000 BOEPD for 10,000-foot laterals -- followed by steep first-year decline.

First-year decline rates in the Delaware Wolfcamp A typically fall in the 70-78% range, measured from peak month production. The second year sees decline moderate to 40-50%, and by year three, wells are declining at 25-35% annually. Terminal decline for mature horizontal Wolfcamp wells clusters around 12.5-14% per year, though a meaningful population of wells decline at rates below 10% or as high as 25% annually.

Recent vintages (2023-2024 completions) show a notable trend: productivity per lateral foot in the Delaware Wolfcamp has been declining. TGS analysis indicates that wells drilled between 2023 and 2024 showed an average 5% productivity decline per lateral foot compared to 2021-2022 completions, with the trend more visible in the Delaware than the Midland side of the basin. This matters enormously for forward-looking type curves. If your 2026 Wolfcamp A type curve is calibrated to 2021 vintage wells, it is likely overstating expected performance by a nontrivial amount.

Wolfcamp A: Midland Basin

The Midland Basin Wolfcamp A behaves differently. IPs are generally lower than the Delaware equivalent, but decline rates tend to be somewhat more moderate, and oil cuts are higher. IP30 rates for 10,000-foot laterals in core Midland County and Martin County areas typically range from 800 to 1,400 BOEPD.

First-year decline rates are slightly less severe than the Delaware -- typically 65-75% from peak month. The flatter decline profile partially offsets lower initial rates, and in many cases Midland Basin Wolfcamp A wells deliver competitive EURs to Delaware wells on a per-foot basis after 5+ years of production.

One development worth watching: Diamondback Energy has reported that early Barnett Shale wells in the Midland Basin track near 75 barrels of oil per lateral foot, compared to roughly 50 bbl/ft for overlying Wolfcamp inventory. If the Deep Barnett emerges as a viable target at scale, it could alter the Midland Basin type curve landscape significantly.

Recent Midland Basin well performance has actually improved slightly in the latest vintages, in contrast to the Delaware's declining trend. This divergence is important for operators and acquirers evaluating relative basin economics.

Wolfcamp B

The Wolfcamp B sits below the Wolfcamp A and is generally a secondary or tertiary target. IPs are 15-30% lower than Wolfcamp A wells in the same area, with more variability. First-year decline rates are comparable (70-75%), but EURs are meaningfully lower, typically 60-75% of a Wolfcamp A type curve in the same section.

The Wolfcamp B is where type curve optimism is most dangerous. Operators frequently apply Wolfcamp A-like economics to Wolfcamp B locations when building inventory counts for investor presentations. In practice, Wolfcamp B wells in Tier 2 and Tier 3 areas often underperform their type curves by 20-30%. Any acquisition target that depends heavily on Wolfcamp B inventory deserves additional scrutiny.

Bone Spring

The Bone Spring formations (First, Second, and Third Bone Spring) in the Delaware Basin present a distinct decline profile. Initial rates are generally lower than Wolfcamp A, but the Bone Spring's shallower depth and lower well costs have historically made it an attractive target.

The concerning trend: Bone Spring productivity has declined more sharply than any other major Permian formation. TGS data shows a 12% decline in productivity per lateral foot for 2023-2024 completions compared to 2021-2022 vintages. Breakeven prices for Bone Spring wells have risen by approximately 16% over the same period.

First-year decline rates in the Bone Spring are typically 65-75%, comparable to the Wolfcamp, but the lower absolute production rate means the economics are more sensitive to both well cost and commodity price assumptions. Bone Spring type curves built on 2020-2021 vintage wells may overstate 2026 expected performance by 10-15%.

How Completion Evolution Has Changed Decline Behavior

It is impossible to discuss Permian decline benchmarks without addressing how completion practices have shifted the curves. From 2018 to 2024, the industry saw several changes:

Lateral lengths increased. Average completed lateral lengths have grown from roughly 7,500 feet to over 10,000 feet, with Permian Resources targeting 10,000-foot averages in 2025 at $775/lateral foot. Longer laterals produce more total volume but decline at similar or slightly higher rates in percentage terms.

Proppant intensity increased, then optimized. The trend from 2018 to 2022 was more sand per foot -- 2,000+ lbs/ft became standard. More recent data suggests diminishing returns above certain thresholds, and some operators have begun optimizing down. Higher proppant intensity tends to increase IP but may not proportionally increase EUR, meaning it can steepen the early decline profile.

Tighter spacing degraded child well performance. In the Delaware Basin, where spacing tightened rapidly through 2017-2018, productivity per lateral foot declined 6% in 2017 and 10% in 2018 relative to 2016 baselines. Parent-child well interference at 660-foot spacing has been documented to reduce 5-year EUR by over 20%. At 750-900 foot spacing, degradation is minimal. The industry has largely corrected to wider spacing, but legacy tight-spacing developments still show up in production databases and can distort basin-level type curve statistics.

The net effect: a type curve built from a heterogeneous dataset of wells spanning multiple completion vintages, lateral lengths, and spacing configurations will not reliably predict the performance of wells being drilled today. Vintage-specific, normalized type curves are essential.

Type Curve Accuracy Analysis

How Operator Type Curves Compare to Actual Production

When public operators disclose both type curves and actual well performance, a consistent pattern emerges: type curves tend to be optimistic for the first 12-18 months and then converge or understate late-life production.

This is not a conspiracy. The explanation is structural. Type curves are typically calibrated to the mean or P50 of a well population, but development programs preferentially drill the best locations first. As an operator works through its inventory from Tier 1 to Tier 2 acreage, the average well quality declines while the type curve remains static. The type curve was accurate for the wells it was built from; it becomes inaccurate for the wells it is applied to.

A second factor: operators often publish "corporate" type curves that blend multiple formations and areas. A corporate Wolfcamp type curve that averages Tier 1 Delaware Basin wells with Tier 2 Midland Basin wells will overstate the expected performance of the Tier 2 locations and understate the Tier 1 locations. For capital allocation, this averaging effect destroys information.

Common Sources of Type Curve Optimism

Survivorship bias in the calibration dataset. Wells that are shut in, workover candidates, or mechanical failures are often excluded from type curve calibration. The remaining "clean" dataset overstates the population average.

Incomplete normalization for lateral length. Operators increasingly report type curves on a per-well basis for 10,000-foot laterals. If the calibration dataset includes 7,500-foot and 8,000-foot laterals that were normalized to 10,000-foot equivalents using linear scaling, the type curve is likely overstated. Production does not scale linearly with lateral length; toe-end stages consistently underperform heel-end stages due to pressure interference and clean-out inefficiency.

Static formation quality assumptions. A type curve calibrated to wells drilled in 2020-2021, when operators were still high-grading Tier 1 inventory, will overstate performance when applied to 2025-2026 locations in less productive rock.

Underestimating the impact of spacing and depletion. In areas with existing production, new wells drilled into partially depleted reservoir will underperform offset type curves built from primary development wells. Stress depletion effects change fracture geometry and reduce stimulated reservoir volume.

Common Sources of Type Curve Pessimism

Overstating terminal decline. Some conservative type curves assume 8-10% terminal decline, but public data shows that a significant population of Permian wells achieve terminal decline rates below 10%, with the most common value around 14%. If you apply a 14% terminal decline to a well that will actually stabilize at 8%, you undercount 5-10% of EUR.

Ignoring completion vintage effects. Older wells in a dataset may pull the average down if completion technology has genuinely improved. Modern Permian completions with optimized cluster spacing and diverter technology can outperform 2018-era completions in the same rock by 10-20%.

Excessive downspacing penalties. After the industry's overcorrection on tight spacing, some operators now apply blanket 15-20% child well penalties even in areas where wider spacing or simultaneous development (wine-rack patterns) mitigates most interference effects.

Tier 1 vs. Tier 2/Tier 3 Inventory: How Much Worse Are the Tails?

This is the question that matters most for forward-looking valuation, and the honest answer is uncomfortable.

Analysis of public production data consistently shows that Tier 2 locations produce 20-35% less EUR per lateral foot than Tier 1 locations in the same formation, and Tier 3 locations are 35-50% lower. The productivity gap compounds with well cost: Tier 2 and Tier 3 locations often require equivalent or even higher well costs (deeper targets, more complex surface logistics, higher water handling) while delivering less production.

The practical implication: an operator or acquisition target with a stated inventory of 500 "Wolfcamp A" locations may have 150 true Tier 1 locations, 200 Tier 2 locations, and 150 Tier 3 locations. If all 500 are assigned the same type curve, the aggregate portfolio EUR is overstated by 15-25%.

With approximately 60-65% of Permian Tier 1 acreage already developed -- a depletion level comparable to where the Eagle Ford and Bakken were in 2018 when production growth stalled -- the composition of remaining inventory skews increasingly toward Tier 2 and Tier 3. Any type curve or valuation that does not explicitly account for this quality degradation is using 2020 assumptions in a 2026 basin.

Decline Curve Methods: What Works in the Permian

Not all decline curve methods are created equal, and the choice of method systematically biases the resulting EUR estimate. Here is how the major approaches compare for Permian tight oil wells.

Arps Decline (Exponential, Hyperbolic, Harmonic)

The Arps family of equations has been the industry standard since 1945. The hyperbolic decline model -- characterized by the initial decline rate (Di), the hyperbolic exponent (b), and production rate (qi) -- remains the most commonly used approach in reserves estimation.

For Permian horizontal wells, the challenge with Arps is the b-factor. Theoretical Arps analysis assumes b values between 0 (exponential) and 1 (harmonic), derived from assumptions of boundary-dominated flow. Permian tight oil wells routinely exhibit apparent b-factors of 1.2 to 2.0+ during extended transient flow, which if extrapolated with Arps hyperbolic decline, produces physically unreasonable EUR estimates -- wells that never stop producing at meaningful rates.

Modified Hyperbolic (b-Factor Switching)

The industry's practical fix: fit a hyperbolic decline with the observed high b-factor during early time, then switch to exponential decline (b=0) at a predetermined terminal decline rate, typically 8-14% per year.

This is the workhorse method for Permian reserves booking. The key engineering judgment is when to switch. A switch at 10% annual decline produces a meaningfully lower EUR than a switch at 6%. For Wolfcamp wells in the Permian, public data suggests the most common terminal decline rate is approximately 14%, though the range spans from below 10% to 25%. The choice of switch point should be informed by actual long-life well data in the same formation and area, not an arbitrary assumption.

Modified hyperbolic works well enough for proved developed producing reserves where you have 2+ years of production history. It is less reliable for PUD reserves where you are projecting from a type curve with no actual well data.

Stretched Exponential Production Decline (SEPD)

The SEPD model, introduced by Valko and Lee, uses a stretched exponential function that provides more flexibility than Arps for tight oil wells. The SEPD tends to produce more conservative EUR estimates than modified hyperbolic, particularly for wells with less than 18 months of production history.

Research comparing Arps, SEPD, and Duong models using shale well data has consistently found that the SEPD model provides more stable EUR estimates as additional production history is added. Where modified hyperbolic EUR estimates can swing significantly depending on the amount of history used for calibration, SEPD estimates tend to converge more quickly. For early-life wells with less than one year of data, SEPD's conservatism is a feature, not a bug.

Duong Model for Tight Oil

The Duong model was specifically developed for tight oil and shale gas wells exhibiting linear flow, which is the dominant flow regime in Permian horizontals for the first several years of production. The model captures power-law behavior and is particularly effective at matching early transient production.

However, the Duong model has a documented tendency to overpredict reserves. Because it assumes long-term linear flow, it does not naturally transition to boundary-dominated flow, and if the reservoir eventually reaches pseudo-steady state, the Duong extrapolation will overshoot actual EUR. Combined with modified hyperbolic, the Duong model consistently yields the highest EUR predictions among common methods.

For Permian wells, the Duong model is best used as a diagnostic tool -- confirming the presence and duration of linear flow -- rather than as the primary forecasting method for reserves booking.

Physics-Informed Machine Learning

The most promising development in decline curve methodology is the integration of physics-based constraints into machine learning models. Rather than fitting empirical equations or letting a neural network learn patterns without physical guardrails, physics-informed approaches encode material balance, flow regime transitions, and pressure-dependent behavior directly into the model architecture.

For a detailed examination of why pure ML decline curve models fail in production settings and how physics-informed approaches address those failures, see our companion article: Decline Curve AI: Physics-Informed vs. Pure ML Approaches.

The relevance to Permian benchmarking is direct: physics-informed models can account for the changing reservoir conditions (depletion, stress changes, interference) that cause traditional empirical methods to overstate EUR in mature development areas. As Permian operators drill deeper into Tier 2 and Tier 3 inventory and increasingly encounter parent-child interference effects, the limitations of purely empirical decline methods will become more consequential.

What the Data Shows About Declining Tier 1 Inventory

The Permian Basin has consumed approximately 60-65% of its Tier 1 inventory. To put that in context: the Eagle Ford and Bakken hit this same depletion level around 2018, and production growth in both basins essentially stopped within twelve months.

The Permian is not the Eagle Ford. It has more stacked pay zones, a larger areal extent, and operators who have learned from the mistakes made in earlier basins. But the physics of depletion are basin-agnostic. When you exhaust the best rock, productivity per well declines, well costs per BOE increase, and production growth requires more capital per incremental barrel.

Current estimates suggest approximately 52,000 remaining locations with Tier 1 or Tier 2 economics in the Permian. That sounds like a large number, but at current drilling rates of roughly 400-450 wells per month, even that inventory represents only 10-11 years of drilling at current pace -- and a shrinking share of those locations will deliver Tier 1 economics.

The EIA's forecast reflects this reality: Permian production is expected to average 6.56 million bpd in 2026, representing a slight decline from the December 2025 peak. More granular drilling productivity data suggests that growth flattens and edges lower in late 2026, with output slipping from mid-2025 peaks.

For type curve work, the implication is clear: forward-looking type curves must be tiered. A single "Wolfcamp A" type curve applied across an operator's entire inventory is not just imprecise -- it is misleading. The geological and engineering data exist to differentiate locations by expected productivity. The failure to do so is usually not technical; it is organizational or political.

A Benchmarking Framework: How to Validate Your Own Type Curves

Whether you are a reservoir engineer preparing a development plan, a VP Engineering approving a capital budget, or a PE firm evaluating an acquisition, here is a practical framework for benchmarking type curves against public data.

Step 1: Define the Comparison Set

Narrow your comparison to wells that match your target on the parameters that matter most:

• •Formation and landing zone (Wolfcamp A Upper vs. Lower, Bone Spring 2nd vs. 3rd)
• •Sub-basin (Delaware vs. Midland, and ideally county-level)
• •Lateral length window (plus/minus 1,000 feet of your target)
• •Completion vintage (within 2 years of your planned completions)
• •Spacing (comparable well density per section)

A comparison set of 30-50 wells meeting these criteria is a reasonable minimum. Fewer than 20 wells and your statistics are unreliable; more than 200 wells and you may be introducing too much heterogeneity.

Step 2: Normalize Consistently

Normalize all production to per-1,000-lateral-foot basis to remove the lateral length variable. Normalize to calendar days (not producing days) to avoid inflating rates for wells with significant downtime. Apply a consistent lookback adjustment to state regulatory data to account for reporting lag.

Step 3: Build P10/P50/P90 Distributions

Do not benchmark against averages. Build cumulative production distributions at 6, 12, 24, and 36 months. Your type curve should correspond to a specific probability level. For SEC reserves booking, the type curve should approximate the P90 for proved undeveloped reserves and no worse than P50 for proved developed. For economic evaluation, the P50 is standard.

If your operator's published type curve exceeds the P50 of your independently constructed comparison set, you have identified a potential source of optimism that warrants further investigation.

Step 4: Compare Decline Rate Trajectories

Plot monthly decline rate (not just cumulative production) for your type curve against the comparison set median. Look specifically for divergence in three windows:

• •Months 1-6: Does the type curve IP match actual wells? If the type curve peak is higher than the P50 of actual wells, the early cash flow is overstated.
• •Months 6-24: Is the type curve decline rate consistent with observed decline? This is where most type curve errors are largest.
• •Months 36+: Does the type curve terminal decline assumption match what long-life wells in the area actually show? Check against the 12.5-14% range commonly observed in Permian Wolfcamp wells.

Step 5: Stress-Test for Tier Quality

If the locations being valued include a mix of Tier 1 and Tier 2 acreage, build separate type curves for each tier. Compare the resulting blended economics against the single-type-curve economics. The difference is the "Tier quality risk" embedded in the valuation.

Step 6: Test Sensitivity to Decline Method

Run the same production dataset through modified hyperbolic, SEPD, and Duong forecasts. If the EUR spread across methods exceeds 15%, the forecast is highly sensitive to methodology choice, which means the production history is likely insufficient to constrain the long-term forecast. In this case, you should weight the more conservative method (typically SEPD) or collect more data before committing capital.

Implications for A&D

The Permian A&D market in 2026 is defined by a paradox: deal flow is robust (U.S. upstream M&A reached $65 billion in 2025 and remains active into 2026), but the quality of available assets is declining along with Tier 1 inventory depletion.

For buyers, the benchmarking framework above is not optional -- it is essential due diligence. Specific recommendations:

• •Never accept the seller's type curve at face value. Rebuild it from public data using the framework described above.
• •Pay particular attention to the Tier 1/Tier 2 split in the stated inventory. A deal that is 70% Tier 2 and Tier 3 locations should be priced accordingly, regardless of what the seller's blended type curve implies.
• •Model decline sensitivity. If the deal economics only work with the Duong model (the most optimistic) and not with SEPD (the most conservative), the valuation is fragile.
• •Watch for spacing-driven degradation risk. If the asset has existing production at tight spacing, new infill locations will likely underperform offset primary wells.

For sellers, intellectual honesty about type curves accelerates deal execution. Sophisticated buyers (and they are all sophisticated in this market) will rebuild your type curves anyway. Overstated type curves do not increase the sale price; they increase the time to close as buyers negotiate the delta during due diligence, or they kill the deal entirely when the buyer's independent analysis diverges too far from the offering memorandum.

For PE firms evaluating E&P acquisitions, the type curve is where most value destruction occurs. The solution is not more complex financial models -- it is better production engineering. Investing in independent decline curve analysis at the individual well level, rather than relying on portfolio-level type curves from the seller's engineering firm, pays for itself many times over in avoided overpayment.

Conclusion

The Permian Basin remains the most important oil-producing region in the Western Hemisphere, but it is no longer a basin where production growth is easy or cheap. Tier 1 inventory depletion, declining well productivity in key formations, and increasingly complex development (parent-child interference, stacked pay development, longer laterals) all demand more rigorous type curve work than was necessary five years ago.

The public data is available to anyone willing to do the work. State regulatory databases, commercial platforms, operator disclosures, and published research provide enough information to build independent benchmarks for any formation and area in the basin. The benchmarking framework in this article is a starting point, not an endpoint -- the specifics of your asset or acquisition target will dictate which formations, vintages, and comparables matter most.

What the data consistently shows is this: the era of easy type curves in the Permian is over. The wells being drilled today are going into different rock than the wells drilled in 2020. The type curves should reflect that reality. When they do not, someone -- the buyer, the lender, or the shareholder -- ends up holding the difference.