Open Tools · Drawcone
Drawcone user guide
Use Drawcone for well drawdown, cone of depression, Theis and Cooper-Jacob pumping-test fits, recovery and a steady capture boundary: inputs, files and stated limits. To use the tool itself, open Drawcone.
Professional use. Results are for education and preliminary design. They must be checked by a qualified professional, and they are not a substitute for code-compliant design or licensed engineering judgement.
Drawcone calculates how the water level around a pumping well falls in an ideal confined aquifer, estimates aquifer transmissivity and storage from a pumping test, and draws the steady capture boundary of one well in a regional flow. Everything runs in your browser; studies save as open JSON and CSV files.
Getting started
Drawcone has three workspaces on the tab row: Response, Pumping test and Capture zone. Everything is entered in SI units: seconds, metres, cubic metres per second and square metres per second. The unit menu in the toolbar changes only how time is shown on charts and tables.
First study in two minutes:
- Open Examples and choose Municipal well, one year of pumping.
- On Response, change the pumping rate in the first row of Pumping history and watch the curves and the cone update.
- Open the Explain a point card to see each term that makes up one drawdown value.
- Use File, Save study to keep the study as a JSON file, or Export on any chart for SVG, PNG or CSV.
Your work is also kept in this browser automatically and restored next time; clear it by opening an example. Undo and Redo step through your edits, and typing in one field counts as one step.
Results are for education and preliminary design. They must be checked by a qualified professional and are not a substitute for code-compliant design or licensed engineering judgement. Drawcone does not establish regulatory capture zones, drinking-water safety or permitting compliance.
Drawdown against time and the cone of depression
Enter the aquifer's transmissivity T (m²/s) and storage coefficient S (a dimensionless fraction). The starting values are illustrative examples, not site data. Replace them with your own values; Drawcone has no site-specific property database.
The response uses the Theis solution for a fully penetrating well in a confined, homogeneous, isotropic aquifer of large extent:
s(r, t) = Q / (4π T) × W(u), where u = r² S / (4 T t)
where W(u) is the exponential integral. Drawdown s is the fall in head from its starting value, so extraction gives a positive value and injection a negative one.
- Drawdown against time shows one curve per observation radius (up to 4) on a logarithmic time axis, with drawdown increasing downward as is usual for water levels. Dotted vertical lines mark changes of pumping rate.
- Cone of depression shows the drawdown out to a radius you choose, at up to 4 elapsed times, mirrored about the well.
The curves are computed at points spaced evenly on a log axis (8 to 400 per curve). Each pumping-rate change adds one term. A term beyond the numeric window of the well function (700 is the largest u it accepts) is dropped only when a bound proves it is below 1e-300 m, and the other terms are kept. A point is left out, with a note saying how many, only when every term there is that small.
Pumping history, injection and recovery
The Pumping history table lists up to 32 rows. Each row is the absolute rate from its start time (seconds); the rate is zero before the first row. Add a row with rate 0 to stop pumping and see the water level recover. Start times must increase. A negative rate is injection.
Drawcone adds the effect of each change of rate by superposition, so recovery is the sum of the continuing pumping term and the term for the shutdown. Where terms nearly cancel to nothing, the engine refuses the result instead of showing an unreliable small value.
The Explain a point card lists every term for one radius and time (at most 64 points are explained at once): the step start, the change of rate, the elapsed time since that step, u, W(u) and the term's contribution. The terms add up to the drawdown shown.
Fit a pumping test
On Pumping test, enter the constant pumping rate, then the observations: time since pumping began (s), radius of the observation well (m) and drawdown (m). Add rows by hand, paste three columns from a spreadsheet, or import a CSV file with the header time_s,radius_m,drawdown_m. Observations from different wells can be mixed because the Theis curve depends on t/r².
Untick a row, or use the Window from row ... to row ... tool, to leave observations out of the fit window. They stay in the file. Drawcone never drops a point on its own, and you choose the window.
Two methods are available:
- Cooper–Jacob (straight line) fits a least-squares line of drawdown against ln(t/r²) and reads T and S from its slope and intercept. It is valid only at late time: every point in the window must have u of 0.01 or less. If a point does not, the fit is refused and says so.
- Theis (curve) fits the full curve. You give a lower and an upper bound for T and for S, and the fit finds the best pair inside them by a bounded local search (the amplitude is solved exactly at each trial). It is not a certified global optimum. Each observation has equal weight in drawdown.
Read the result carefully: the fitted T and S, the residual RMS in metres, the relative residual (RMS divided by the largest observed drawdown) and the degrees of freedom. If the fit did not converge, or if the best fit touches a bound you set, Drawcone says so; do not accept those parameters. Drawcone gives no confidence intervals or p-values. A small residual does not prove that the aquifer behaves like the model, so also read the residual chart for trends. Changing any input clears the fit result; run it again.
Capture zone of one well
Enter the pumping rate Q, the transmissivity T and the regional hydraulic gradient i. The well is at the origin and the regional flow runs in the +x direction. The steady boundary that divides water that reaches the well from water that passes it follows from a stream function: the stagnation point lies down-gradient at a = Q / (2π T i), and far upstream the zone approaches the width W = Q / (T i). The flux balance shown confirms that the pumped water equals the regional flow across that width.
The Plan view shows a labelled near-well window with equal physical axis scales and a full-sample overview with independently scaled axes. The overview includes every computed coordinate; it is not the end of the zone. The zone is open upstream. The boundary approaches the half-width lines but never closes, and the last plotted point is not the limit of the zone. Never close the coordinates into a protection-zone polygon. This model has one well, uniform steady flow and no travel-time contours, porosity, leakage, dispersion, boundaries or neighbouring wells, and storage does not enter a steady boundary. It is a screening picture, not a delineation.
Save, Open, import and export
Save writes an open, versioned JSON file with the aquifer, pumping history, the points requested, observations, fit bounds, capture settings, notes and the workspace settings. Saving never needs a calculation to have run. Open reads the same format strictly: an unknown field, a wrong type or a changed disclaimer is refused with a message, never silently repaired. Files up to 8 MiB are accepted.
Observation CSV files have exactly the columns time_s,radius_m,drawdown_m in SI units; units are not guessed from names, blank or formula cells are refused, and the file may carry comment lines starting with #. Exports place the attribution and the professional-check notice in comment lines, quote the study title and neutralise spreadsheet formulas. Chart exports as SVG or PNG carry a footer with the title and the date in Mountain Time.
Only the documented JSON and CSV formats are supported. There is no import of other programs' project files.
Assumptions and limitations
The models assume a homogeneous, isotropic, confined aquifer of infinite extent with constant properties, a fully penetrating well of negligible radius and instantaneous release of water from storage. They leave out leakage, delayed yield and unconfined behaviour, boundaries, partial penetration, well-bore storage and skin. Do not infer that these assumptions hold from a small residual alone.
- At most 10,000 observations or response points, 32 pumping steps, 128 fit iterations and 1,001 points on each capture branch.
- Transmissivity, times and radii must lie between 1e-12 and 1e12 in SI units; storage between 1e-12 and 1; the gradient above 0 and at most 1.
- The well function is evaluated for 1e-300 ≤ u ≤ 700; outside that range, or where a result overflows, underflows or loses all precision, a message replaces the number.
- Each analysis stays under a conservative working-memory estimate of 16 MiB; this is an estimate, not a measured guarantee.
Messages and what to do
A refused calculation shows a short code and a plain message.
| Code | Meaning | What to do |
|---|---|---|
| INPUT | A value is missing, malformed or outside its window | Fix the highlighted field |
| SIZE | A list or file is too large | Reduce points, steps or rows |
| REGIME | A point is outside the range where a formula or its numerics hold | Use later times, smaller radii or a different window |
| FIT | The data cannot identify T and S, or the straight-line slope is not positive | Add observations that span more of t/r², or change the window |
| NUMERIC | A result would overflow, underflow or cancel to nothing | Use less extreme values |
| FILE | A file is not a valid Drawcone document | Check that it came from Drawcone and is unedited |
| GATE | The tool could not confirm it is running on the site | Reconnect and try again |
How we verify Drawcone
The Verify page runs 24 reference cases in 7 families in your browser, through the same engine the tool uses, and shows each computed value, the reference value and the allowed difference. A "verified" badge appears only after a complete run in which every case passes.
- The well function and the Theis response are compared with an independent 70-digit numerical quadrature, not with the series the tool uses.
- A finite pumping period followed by recovery is compared with a direct integral of the pumping history.
- The capture width is checked by flux conservation, and the boundary is checked to be a streamline.
- The fits recover known parameters from data generated independently, and the straight-line fit is checked algebraically.
- A published worked example: the drawdown table of one observation well in a classic United States Geological Survey report on ground-water hydraulics (Professional Paper 708, 1972, Table 6, well N-1, 25 readings), fitted here to T and S. The report quotes T of about 13,700 ft²/day and S of about 2e-4 from a graphical curve match, The report prints no accuracy for that match, so this case uses project-selected tolerances of 6 percent on T and 15 percent on S that cover the printed rounding; a separate test checks the fit tightly against an independent least-squares calculation. It is a textbook example, not a field validation.
A complete published field pumping test with checked units is still pending, is shown as pending, and is never counted as a pass. These checks show that the implementation matches the ideal models. They do not show that a real aquifer behaves like them. Sources opened on 2026-10-26: S. W. Lohman, Ground-Water Hydraulics, U.S. Geological Survey Professional Paper 708 (1972); NIST Digital Library of Mathematical Functions, section 6.6. Formulas are implemented originally; no source tables or figures are reproduced beyond the one cited data table of readings.
When to use something else
Drawcone is a compact analytical tool. For leaky, unconfined or fractured aquifers, wellbore effects, travel-time capture zones, several interacting wells, boundaries or a meshed regional model, use a dedicated groundwater package or a numerical model. For seepage through a meshed domain, see Meshfield in the Open Tools.