Drawing No. EH–FM–006 // Fluid Mechanics & Piping
Excavation Dewatering Screening Model
Reviewed August 2026
Get an order-of-magnitude estimate of groundwater inflow to an excavation, using Sichardt's empirical radius-of-influence formula together with the classical Dupuit or Thiem equations for radial flow to a well.
What problem does this solve?
Before an excavation can be dug below the water table, the design team needs a rough idea of how much water will need to be pumped to keep it dry, and how far the resulting drawdown might extend (relevant for assessing settlement risk to nearby structures and impacts on neighboring wells). This screening model gives that first-pass estimate using well-established empirical and analytical relationships, appropriate for early planning before a full hydrogeological investigation.
Inputs
Results
Background
R = 3000·s·√K, where R is the radius of influence (m), s is drawdown at the excavation/well face (m), and K is hydraulic conductivity (m/s) — note the formula is dimensionally inconsistent (it hides an implicit unit conversion) and is purely empirical, first published by Kyrieleis and Sichardt in 1930. Despite well-documented accuracy limitations, it remains the most widely used rule-of-thumb radius-of-influence estimate in geotechnical practice because of its simplicity.
Q = πK(H² − h²)/ln(R/r), where H is the initial saturated thickness, h is the depth of saturated flow remaining at the excavation face (H minus drawdown), R is radius of influence, and r is the excavation's equivalent radius. This treats the excavation as a large-diameter well and assumes steady-state, radially symmetric flow with the classic Dupuit-Forchheimer assumption of horizontal, depth-averaged flow.
Q = 2πKb(H − h)/ln(R/r), where b is the confined aquifer's constant saturated thickness and (H − h) is the drawdown. Unlike the unconfined case, the aquifer's transmissive thickness stays constant since it's bounded above by an impermeable confining layer rather than by the free water table itself.
Real excavations are rectangular, not circular, but the Dupuit/Thiem equations assume radial flow to a circular source. A common approximation converts a rectangular excavation of length L and width W to an equivalent circular radius using r ≈ √(LW/π) (equal-area equivalent), which this calculator expects as a direct input — convert your actual excavation footprint before entering it.
Both Sichardt's formula and the Dupuit/Thiem equations assume a homogeneous, isotropic aquifer of effectively infinite lateral extent, fully penetrating wells, and steady-state flow — conditions real sites rarely meet exactly. Layered soils, nearby recharge boundaries (rivers, other water bodies), partial penetration, and transient (time-varying) flow all change real inflow significantly from this idealized estimate. A qualified hydrogeologist typically confirms or refines these numbers using site-specific pumping tests before final dewatering system design.
Frequently asked questions
Practical questions about inputs, assumptions and interpretation.
Independent studies comparing it against more rigorous transient analytical solutions (such as the Theis or de Glee equations) have found it can differ by a factor of two or more depending on conditions, and it doesn't account for time, aquifer storage properties, or boundary conditions that a full transient analysis would capture. It endures mainly because it needs only two easily estimated inputs (drawdown and hydraulic conductivity) and gives a defensible order-of-magnitude starting point for early planning.
In an unconfined aquifer, drawing down the water table also reduces the aquifer's own saturated thickness near the well, which itself reduces the aquifer's transmissive capacity there — captured by the H² − h² term. In a confined aquifer, the aquifer stays fully saturated (bounded by an impermeable layer above) regardless of how much the piezometric head drops, so only the linear (H − h) drawdown term appears, without the same self-limiting effect.
That's not physically valid input for the Dupuit unconfined equation — you can't have negative remaining saturated thickness at the well face. This usually signals the excavation needs to fully dewater an unconfined layer, which requires a different analysis (often involves the aquifer effectively going dry near the excavation, needing numerical or more advanced analytical treatment) rather than this simplified formula.
The steady-state inflow rate from this screening estimate gives the total pumping capacity a dewatering system needs to provide, which then informs how many wellpoints or deep wells are needed (each with its own realistic per-well capacity) and their spacing — but detailed system layout, well screen design and pump selection require more detailed analysis beyond this screening-level total inflow number.