Decline Curve Software
Arps every well. Re-fit every shift.
Most operators run decline curve analysis quarterly, in a reservoir engineering review. By then three months of underperformance has already happened. Continuous decline monitoring (forecast updated nightly, actuals checked against confidence bands every shift) turns the production accounting report into a same-day intervention decision. Slide the inputs below to see the math change in real time.
Interactive demo
Drag the Arps inputs. Watch the curve respond.
Initial rate, decline coefficient, b-factor. Same equation a reservoir engineer uses. Same equation WorkSync runs nightly against every well.
Arps regimes
Three forms. When each one fits.
Exponential (b = 0)
q = qi × exp(-D × t)
When: Terminal decline, mature conventional wells, late-life unconventional
Constant fractional decline rate. Simplest. Most defensive reserves case.
Hyperbolic (0 < b < 1)
q = qi / (1 + b × D × t)^(1/b)
When: First 12-18 months of unconventional wells (Permian, Bakken, Eagle Ford, etc.)
b typically 0.4-0.6 for tight oil. Captures the early-life flattening.
Harmonic (b = 1)
q = qi / (1 + D × t)
When: Edge case. Most common in solution-gas-drive reservoirs at low pressure
Slowest decline shape. Rare in modern unconventional fits.
The pipeline
Five stages from raw rate to ranked work plan.
Stage 01
Per-well fit
Each well gets an Arps fit against historical production. b-factor selected based on play, completion era, and operating regime. qi and D fit to actual rate history.
Stage 02
Continuous re-fit
Every shift, new production data lands. The fit is updated. Regime-change detection flags wells transitioning from hyperbolic to exponential (a different b-factor takes over).
Stage 03
Confidence band
A ±8 BO/d (or play-appropriate) confidence band wraps the forecast. Normal variability stays inside. Crossings get flagged.
Stage 04
Deficit attribution
When actual drops below the lower band, the deviation is attributed: mechanical (rod-pump fillage), reservoir (pressure depletion), facility (compression / choke), data (SCADA gap). Operator gets a categorized work order, not just an alarm.
Stage 05
Ranked work plan
Categorized deviations join the Work Engine's ranked daily plan with dollar-impact-of-waiting attached. The reservoir engineer's analysis becomes the foreman's 6 AM work order.
Continue the cluster
Decline curve analysis is the baseline. Anomaly detection is the trigger.
Companion pillars
The cluster
How the loop closes
Capability deep-dives
By basin
Per-basin decline character
Frequently asked
What reservoir engineers ask about continuous decline.
What is decline curve software?
Software that fits a mathematical decline model (Arps exponential / hyperbolic / harmonic) to each well's historical production rate and projects future production. Foundation of reservoir engineering, reserves estimation, and short-term operational decisions.
What is the Arps equation?
Three forms. Exponential: q = qi × exp(-D×t). Hyperbolic: q = qi / (1 + b×D×t)^(1/b). Harmonic: q = qi / (1 + D×t). Most US onshore horizontals are hyperbolic with b ≈ 0.5 for first 12-18 months, then exponential terminal decline.
Continuous vs quarterly review?
Quarterly review catches a well that has been underperforming for 90 days. Continuous monitoring catches it in 24-72 hours. Same math, different cadence. The actionable window goes from "next slide deck" to "Tuesday's ranked plan."
How is AI-assisted decline different from traditional fitting?
Continuous re-fit as new data lands. Regime-change detection (hyperbolic → exponential transition). Event-driven adjustments (workover, choke change, shut-in). Underperformance attributed to mechanical, reservoir, or facility categories. The reservoir engineer keeps the final call.
How does it integrate with anomaly detection?
Arps is the baseline. Per-well ML anomaly detection learns the normal variability around the curve and flags deviations beyond expected noise. Arps gives long-term shape; ML catches short-term break.
What is the ROI?
Catch-time: a well flagged in 24-72 hours instead of 30 days costs 5-15% to remediate vs after compounded deferred production. Better reserve estimates: continuously calibrated curves produce more defensible 1P/2P/3P bookings.
See continuous decline running on your wells.
4-week pilot at no license cost. Pick one field. Day 7 you see the forecasts. Day 14 you see the deficit attribution. Day 28 you decide.