Decline Curve Analysis: Oil Value

Table of Contents

    Decline curve analysis connects an oil well’s production history to a forecast of future output. For an investor, the important questions are how many saleable barrels remain, when they will be produced, and whether their cash flows justify the price paid for the asset.

    A producer can report growing total output while its existing wells decline. In a November 2025 analysis, the U.S. Energy Information Administration reported that Lower-48 oil production from wells that started producing in 2023 or earlier fell from 11.0 million barrels per day in December 2023 to 6.7 million in December 2024. New wells offset those declines. That historical cohort is not a decline curve for an individual well, but it explains why a company forecast must account for both existing production and investment in new production. [1]

    This guide compares exponential, hyperbolic and harmonic curves, integrates one hypothetical oil forecast, and carries the resulting annual barrels into a cash-flow valuation. The mathematics is general; the investment discussion assumes a U.S. reader. Every worked production, cost and price input below is hypothetical, rather than a reserve determination or an engineering recommendation.

    What Does Decline Curve Analysis Forecast?

    A Rate, a Volume and a Value

    A decline curve forecasts a production rate, such as barrels per day. Integrating that rate over a defined future operating life gives remaining production. Applying ownership, prices, costs and discounting gives a cash-flow value. Each step needs additional assumptions.

    The classical Arps family extrapolates a declining production trend using a starting rate, an initial decline rate and a shape parameter. It is an empirical description of performance, so a close historical fit does not establish that the same trend will persist. ResInsight’s developer documentation presents the three familiar forms and their rate-cumulative relationships. [2]

    Choose the forecast origin carefully. Here, t = 0 is the valuation date, and qi is the production rate at that date. It need not be the well’s first-ever or peak production rate. Production already recovered before that date belongs in historical cumulative production, not in the future cash-flow forecast.

    Keep the reporting question separate. A curve calculation alone does not establish proved reserves, development readiness or a commercial project. The guide to oil reserves and valuation covers reserve classifications and discounted reserve disclosures.

    Exponential vs Hyperbolic vs Harmonic Decline

    Let q(t) be the forecast production rate, Di the initial instantaneous nominal decline rate, and b the dimensionless decline exponent. In this article, t is measured in years, Di in inverse years, and q in gross stock-tank barrels per calendar day. A stock-tank barrel measures oil volume at the stated surface conditions; it is not a barrel of oil equivalent combining different products.

    Instantaneous Nominal Decline
    D(t) = −(1 / q(t)) × dq(t)/dt
    D(t) = Di / (1 + bDit)
    D(t) is positive when production falls. Exponential decline keeps it constant; a positive b makes it decrease over time.
    The three classical decline-curve forms
    Model b convention Production rate q(t)
    Exponential b = 0 qi × e−Dit
    Hyperbolic 0 < b < 1 qi / (1 + bDit)1/b
    Harmonic b = 1 qi / (1 + Dit)

    Exponential and harmonic are the limiting cases of the family. With the same qi and Di, a larger b produces a slower subsequent decline and a larger late-life production tail. These equations and the conventional classical range follow the readable rate-time formulas in ResInsight. [2]

    “Hyperbolic” is also used for empirical fits with b above one, particularly when analysts model early unconventional production. That usage does not make an unrestricted high-b extrapolation a reliable long-term recovery estimate. The applicable flow regime and terminal assumptions still need support.

    Nominal vs Effective Decline: Define the Period

    A “50% annual decline” is ambiguous until the convention is specified. Nominal decline here means the instantaneous fractional rate of production decline. Effective decline means the fraction by which the production rate falls across a stated interval. These production terms are separate from nominal versus inflation-adjusted dollars in valuation. Harmony’s documentation makes this instantaneous-versus-period distinction explicit. [3]

    Effective Decline Over an Interval
    deff(t, Δt) = 1 − q(t + Δt) / q(t)
    Both rates must use the same production basis. State the start date and interval length.

    For exponential decline, deff = 1 − e−DiΔt. If Di = 0.50 per year and the interval is one year, effective decline is 1 − e−0.50 = 39.35%. A 50% effective annual exponential decline instead requires Di = −ln(1 − 0.50) = 0.693147 per year.

    For the first year of a hyperbolic or harmonic forecast, divide its year-one rate by qi. Hold qi = 100 bbl/day and initial nominal Di = 0.50 per year fixed:

    Equal initial nominal decline, different first-year changes (hypothetical)
    Curve First-year effective decline Year-one rate
    Exponential, b = 0 1 − e−0.50 = 39.35% 60.65 bbl/day
    Hyperbolic, b = 0.5 1 − 1.25−2 = 36.00% 64.00 bbl/day
    Harmonic, b = 1 1 − 1.50−1 = 33.33% 66.67 bbl/day

    The equal initial nominal input does not hold the first-year ending rate fixed. For positive b, effective decline over a later one-year interval also changes as the well ages. An exponential conversion should not be applied unchanged to every hyperbolic period.

    From Production Rate to Cumulative Volume

    Production over a year is the area under the rate curve during that year. Multiplying a declining well’s ending daily rate by the number of days understates that year’s output, because earlier daily rates were higher.

    Future Cumulative Production Through T Years
    Q(T) = 365 × ∫0T q(t) dt
    Period output = Q(tend) − Q(tstart)
    Q is in barrels when q is in barrels/day and t in years, using exactly 365 days per year for these hypothetical calculations.

    Integrating the three rate equations gives the following finite-horizon quantities. They are the rate-cumulative relationships expressed using elapsed time, with the daily-to-year conversion made explicit. [2]

    Exponential: b = 0
    Q(T) = (365qi / Di) × [1 − e−DiT]
    Hyperbolic: 0 < b < 1
    Q(T) = {365qi / [Di(1 − b)]} × [1 − (1 + bDiT)(b−1)/b]
    Harmonic: b = 1
    Q(T) = (365qi / Di) × ln(1 + DiT)

    Use actual calendar conventions in a real model. Monthly volumes, calendar-day rates and producing-day rates cannot be substituted for one another without adjustment. As a dated example, EIA’s discontinued AEO2022 methodology omitted a potentially partial first production month and normalized other months to 30.4 days. Its March 2023 description also used a specified hyperbolic-to-exponential transition and a 30-year total horizon. Those were that model’s assumptions, rather than universal rules for well life or recovery. [5]

    A Worked Remaining-Production and EUR Example

    Continue with a hypothetical existing oil well at the valuation date. Set qi = 100 gross bbl/day, Di = 0.50 per year nominal, and b = 0.5. Assume the forecast already includes calendar downtime, so no additional uptime multiplier is needed. Let prior gross cumulative production be 27,000 barrels. First assume production stops at 25 gross bbl/day; the cash-flow section below derives this threshold from the operating-cost assumptions.

    Calculate the Finite Forecast

    The rate equation is q(t) = 100 / (1 + 0.25t)2. At the cutoff, 25 = 100 / (1 + 0.25T)2. Thus 1 + 0.25T = 2, giving T = 4 years.

    The cumulative equation simplifies to Q(T) = 146,000 × [1 − 1 / (1 + 0.25T)]. Future production is Q(4) = 146,000 × (1 − 1/2) = 73,000 gross barrels.

    Scenario estimated ultimate recovery is 27,000 prior barrels + 73,000 future barrels = 100,000 gross barrels.

    Under PRMS terminology, EUR includes prior production and is not itself a reserve class. “73,000 remaining barrels” and “100,000-barrel EUR” therefore describe different quantities. Both use the same gross ownership and oil-volume basis here; neither has been certified as proved reserves. [4]

    To build annual revenue, take differences between cumulative values. For example, Q(1) = 146,000 × (1 − 1/1.25) = 29,200 barrels; year two contributes Q(2) − Q(1) = 48,666.67 − 29,200 = 19,466.67 barrels, rounded.

    Hypothetical production schedule, using unrounded calculations
    Year-end Ending rate (gross bbl/day) Output during year (gross bbl) Future cumulative output (gross bbl)
    1 64.00 29,200.00 29,200.00
    2 44.44 19,466.67 48,666.67
    3 32.65 13,904.76 62,571.43
    4 25.00 10,428.57 73,000.00

    Year four produces 10,428.57 barrels, rather than 25 × 365 = 9,125 barrels. The latter uses only the final day’s rate. The cumulative column starts at the valuation date; adding the prior 27,000 barrels to it would describe lifetime cumulative production.

    Terminal Decline Is Different From the Final Cutoff

    A terminal exponential decline changes the late-life shape of the forecast. The model follows a hyperbolic curve until its instantaneous nominal decline reaches a chosen terminal rate, then continues exponentially from the production rate at that transition. Matching the rate and decline convention avoids an artificial jump. It does not mean production stops at the transition. Harmony describes this modification as a response to excessive long-term hyperbolic extrapolation. [3]

    The final cutoff stops the production forecast. It may be imposed by economics, technical limitations or the end of operating rights. PRMS truncates recovery at the earliest applicable limit. [4] A terminal exponential phase cannot justify production beyond an earlier license or economic endpoint.

    Do Not Treat an Infinite Tail as Recoverable Oil

    The pure harmonic curve’s cumulative volume grows without bound if extrapolated forever. An unrestricted b above one also produces a divergent mathematical integral. Even a curve with a finite integral does not establish that every theoretical barrel can be recovered or sold economically. State the final horizon and any terminal-decline assumption explicitly.

    In the worked hyperbolic example, the infinite-time mathematical total would be 146,000 future barrels, twice the finite 73,000-barrel forecast. The operating cutoff is therefore a material assumption, even though b is below one.

    Turn the Production Forecast Into Oil-Asset Cash Flows

    Return to the same well. Assume a 100% working interest that bears all modeled costs and an 80% net revenue interest after a 20% royalty. Use a constant realized oil sale price of $70 per barrel before royalty, variable cash costs of $20 per gross barrel, and avoidable fixed operating cash costs of $328,500 per year, or $900 per day. The variable cost includes the assumed variable transportation cost.

    The sale price is an asset-specific realized-price assumption. A benchmark alone does not establish what a producer receives; quality, location and contracts matter in crude oil markets. If a well also sells gas or liquids, forecast those products and their costs separately, using the relevant natural gas market assumptions rather than assigning the oil price to every equivalent barrel.

    Derive the Simple Economic Rate

    Contribution after royalty and variable cost is 0.80 × $70 − $20 = $36 per gross barrel. In this constant-price, constant-cost example, ongoing operating cash flow per day is q × $36 − $900. It reaches zero at qe = $900 / $36 = 25 bbl/day, the cutoff used above.

    This simplified rule assumes no temporary shut-in/restart decision, no earlier technical or license limit, and no further development project. PRMS’s economic-limit test considers avoidable operating cash costs and excludes closure costs from that particular calculation. Closure still belongs in the valuation. Temporary negative periods can require a broader cumulative cash-flow assessment rather than immediate permanent abandonment. [4]

    Build the Asset DCF

    Assume the existing asset costs $500,000 at time zero, needs no further development capital, and incurs $100,000 of closure cash cost at the end of year four, with no salvage value. Use a hypothetical 10% annual effective real discount rate and constant-purchasing-power dollars. For this teaching case, omit income and production taxes, corporate overhead, financing, hedges and working-capital changes. Assign each year’s cash flows to year-end, an approximation to receipts and payments throughout the year.

    Simplified Pretax Asset Cash Flow
    CFj = (0.80 × P − $20) × Vj − Fj − Aj
    Acquisition NPV = −$500,000 + Σj CFj / (1 + r)tj
    P is realized USD per barrel; V is gross barrels during the period; F is fixed operating cash cost; A is closure cash cost; t is payment time in years.

    For year one, $36 × 29,200 − $328,500 = $722,700. Multiply gross volume by the revenue interest only in calculating sales; the 100% working interest still bears costs on gross production. Applying the 80% factor to both revenues and costs would overstate this investor’s margin.

    Hypothetical year-end asset cash flows, rounded to USD
    Year Contribution after royalty and variable cost Fixed operating cost Closure cost Net cash flow
    1 $1,051,200 $328,500 $0 $722,700
    2 $700,800 $328,500 $0 $372,300
    3 $500,571 $328,500 $0 $172,071
    4 $375,429 $328,500 $100,000 −$53,071

    Year four has positive operating cash flow before closure, but its final net cash flow is negative because closure is paid then. The rounded rows above are for presentation; discount the unrounded calculations.

    Discount the Cash Flows and Deduct the Purchase Price

    Future cash-flow PV = 722,700 / 1.10 + 372,300 / 1.102 + 172,071.428571 / 1.103 − 53,071.428571 / 1.104 = $1,057,717, rounded.

    Acquisition NPV = $1,057,717 − $500,000 = $557,717, rounded.

    This is an assumption-dependent pretax asset NPV, not a filed PV-10, an after-tax equity value or a recommended bid. A company analysis must reconcile its other assets, corporate cash costs, taxes and financial claims. Use the broader discounted cash flow framework for that task and net present value for the purchase-price decision.

    How b and Oil Prices Change the Valuation

    Change b While Holding the Initial Nominal Decline Fixed

    Hold the preceding price, cost, interest and discount assumptions fixed. Change only b, keeping qi = 100 bbl/day and initial nominal Di = 0.50 per year. Recalculate both the operating life and cash flows, stopping each pure curve when it reaches 25 bbl/day. No separate terminal exponential phase is imposed in this finite comparison.

    Hypothetical b sensitivity at a $70 realized oil price
    Curve Economic life (years) Remaining gross barrels Acquisition NPV (USD)
    Exponential, b = 0 2.773 54,750 $354,448
    Hyperbolic, b = 0.5 4.000 73,000 $557,717
    Harmonic, b = 1 6.000 101,199 $823,832

    The exponential cutoff is ln(100 / 25) / 0.50 = 2.772589 years; the harmonic cutoff is (100 / 25 − 1) / 0.50 = 6 years. The larger b produces more barrels and a higher NPV under these fixed starting assumptions. It also changes the early path, as the first-year effective-decline comparison showed. These are parameter sensitivities, not three equally credible fits to actual well history or low/base/high reserve categories.

    Change Price and Recalculate the Economic Cutoff

    Now return to b = 0.5 and keep the other base assumptions. Test realized prices of $50, $70 and $90 per barrel. For each, calculate qe = 900 / (0.80P − 20). Higher prices support a lower final economic rate; lower prices stop production earlier.

    Hypothetical price scenarios change margin and future economic volume
    Realized USD/bbl Economic cutoff (gross bbl/day) Life (years) Remaining gross barrels Acquisition NPV (USD)
    $50 45.00 1.963 48,060 −$300,149
    $70 25.00 4.000 73,000 $557,717
    $90 17.31 5.615 85,260 $1,596,710

    At $50, the contribution is 0.80 × $50 − $20 = $20 per gross barrel, so qe = 900 / 20 = 45 bbl/day. Solving the same hyperbolic rate equation gives T = [sqrt(100 / 45) − 1] / 0.25 = 1.962848 years. Inserting that T into Q(T) gives 48,060 barrels, rounded. The price decline reduces both cash earned per barrel and the volume produced before the economic cutoff.

    For noninteger lives in both tables, full-year cash flows occur at year-end, and the final partial period ends at the exact cutoff. Its volume is the difference in Q, its fixed cost is prorated by the fraction of a year, and closure is paid once at that cutoff; discounting uses the exact fractional year. The displayed rounded lives are not calculation inputs.

    Discounting is another independent sensitivity. With base production and costs unchanged, acquisition NPV is approximately $585,941 at 8% and $530,812 at 12%, versus $557,717 at 10%. None of these rates is established as the asset’s correct required return. Use sensitivity analysis to expose assumptions, and distinguish a scenario range from a statistical confidence interval.

    How to Evaluate a Decline-Curve Forecast

    1. Reconcile the data. Identify product, gross/net basis, stock-tank or other measurement conditions, producing days, downtime and partial months. Match forecast start with historical cumulative output.
    2. Choose the fitting interval for a reason. Separate ramp-up, shut-ins, workovers and operating changes from the decline segment. Explain why the observed history can support the projected regime.
    3. State every curve convention. Record qi, nominal or effective decline, its time unit, b, terminal exponential assumptions and final technical/license/economic cutoff.
    4. Check out-of-sample plausibility. Compare forecasts with later observed production where available and with relevant engineering evidence. A small historical fitting error alone does not validate a large future tail.
    5. Rebuild cash flows on the investor’s interest. Reconcile royalty-adjusted revenue with cost-bearing ownership, future capital, product-specific realized prices and closure costs. Keep real/nominal and pretax/after-tax assumptions consistent.
    6. Separate existing wells from growth spending. New drilling, acquisitions and workovers can support company output, but their production and cash costs need explicit forecasts. A single existing-well decline cannot stand in for the entire business.

    Supply constraints, construction, operating performance and contractual risks also affect whether forecast barrels become cash. Those are part of commercial risk in project finance; a smooth production curve does not resolve them.

    Common Mistakes in Decline Curve Analysis

    • Mixing nominal and effective decline. Define the interval and convert using the correct curve, rather than treating every quoted percentage as interchangeable.
    • Mixing daily rates and yearly time. Include the day-to-year conversion in volume integration; distinguish operating-day rates from calendar-day rates.
    • Using the ending rate as annual production. Integrate the rate across the year or use cumulative differences.
    • Calling remaining output EUR. Add prior production only when calculating lifetime recovery on the same basis; keep those historical barrels out of future revenue.
    • Choosing b to maximize value. Fit and challenge the production assumptions before relying on the most attractive tail.
    • Keeping economic barrels fixed when prices fall. Reassess the cutoff and recovery as well as the revenue per barrel.
    • Adding a perpetual terminal value to a depleting well. Include only supported remaining production, other assets and costs; do not invent replacement reserves.
    • Equating a positive asset NPV with a stock recommendation. The result depends on the price paid, modeling scope and risks outside this simplified asset.

    Limitations: Flow Regimes, b Fitting and Uncertainty

    Traditional decline extrapolation is most defensible when the analyzed history reflects an appropriate established drainage regime and sufficiently stable operating conditions. Early transient flow, changing flowing pressure, artificial lift, interference and changes in fluid production can break the apparent trend. In very low-permeability reservoirs, the transient period can be prolonged. PRMS specifically cautions that these factors complicate performance forecasting. [4]

    Different combinations of initial rate, initial decline and b can fit a short history while implying very different future output. A fitted b above one may describe a transient segment; carrying it unchanged to remote years can exaggerate recovery. A terminal switch bounds one aspect of that extrapolation, but choosing a switch does not validate the geology or make the remaining forecast certain.

    Oil, gas and water production, operating constraints and depletion mechanisms should inform the engineering assessment. An aggregate company curve also mixes wells of different ages with drilling and acquisitions, so it can conceal existing-well decline. Financial analysis adds further uncertainty through future prices, taxes, costs, capital requirements and discount rates. These equations help make those assumptions visible; they do not replace reservoir evaluation, reserve-reporting requirements or a complete investment analysis.

    Frequently Asked Questions

    b controls how quickly the instantaneous nominal decline itself decreases over time. The classical exponential case has b = 0, hyperbolic lies between zero and one, and harmonic has b = 1. A larger b produces a longer tail when the starting rate and initial nominal decline are held fixed; it is not a recovery probability.

    No. For exponential production, a nominal decline of 0.50 per year means the rate falls by 1 − e−0.50, or about 39.35%, over one year. Hyperbolic effective decline also depends on b and the start of the interval. State which percentage the forecast uses.

    Remaining production starts at the forecast date. EUR adds the oil already produced to the estimated future recovery under the stated conditions. In the example, 73,000 future gross barrels plus 27,000 prior gross barrels gives a 100,000-barrel scenario EUR. A calculated EUR does not by itself establish a proved-reserve classification.

    No. A terminal exponential phase changes the decline law used for later production. The final production cutoff is a separate technical, license or economic endpoint. Define both; a terminal curve cannot extend an operating right or make subeconomic production profitable.

    Lower prices reduce contribution per barrel. With fixed operating costs, production can become uneconomic at a higher daily rate and stop earlier. The physical curve need not change for the economically produced volume to fall. Actual contracts, taxes, cost reductions and shut-in/restart options can modify the result.

    Sources

    1. U.S. Energy Information Administration, Rapid declines from horizontal wells require more drilling to sustain production, November 5, 2025. Historical Lower-48 production by well vintage.
    2. ResInsight developer documentation, Decline Curve Analysis, Origins, Rate-Time and Rate-Cumulative Production sections. Readable equations for the classical family.
    3. IHS/Harmony official documentation, Decline Analysis Theory, Practical Key Points, Nominal vs Effective and Modified Hyperbolic Decline sections.
    4. Society of Petroleum Engineers and sponsoring societies, Petroleum Resources Management System, Revised June 2018 (v. 1.03), sections 1.1, 3.1 and 4.1.4. International definitions, economic limits and production-performance cautions.
    5. U.S. Energy Information Administration, Production Decline Curve Analysis in the Annual Energy Outlook, March 16, 2023. Historical methodology for AEO2022; explicitly discontinued.

    Sources accessed November 15, 2026. Worked examples and sensitivity scenarios are independently calculated hypothetical illustrations, not source-reported well results or commodity-price forecasts.

    Disclaimer

    This article is for educational and informational purposes only and does not constitute investment, legal, tax or engineering advice. Hypothetical decline curves and cash flows simplify actual operating and commercial conditions. They do not certify reserves, determine a property’s fair value or recommend an investment. Review current source data and obtain qualified advice for a specific asset or security.