Suction anchors (floating OWT)

The op3.anchors package extends Op3 from fixed-bottom foundations to floating-platform suction-anchor mooring design. It covers ultimate capacity (four analytical methods + an FE-calibrated path), installation feasibility, optimal padeye depth (including a novel dissipation-centroid method), cyclic-storm degradation, and MoorPy coupling for full safety-factor time-series checks.

Scope and standards

The module is built around the design framework of:

  • DNV-RP-E303 (2021) – Geotechnical Design and Installation of Suction Anchors in Clay (default capacity method).

  • API RP 2SK (2005, reaff. 2015) – Stationkeeping Systems for Floating Structures.

  • ISO 19901-7 (2013) – Stationkeeping systems.

  • DNV-ST-0119 (2021) – Floating wind turbine structures (consequence-class partial safety factors).

It is restricted to clay sites with linearly increasing undrained shear strength \(s_u(z) = s_{u0} + k_z\,z\). Layered profiles require segment-wise integration; sand sites are out of scope (no analytical Np table is published for granular anchors).

Quick start

from op3.anchors import (
    SuctionAnchor, UndrainedClayProfile, MooringLoad,
    anchor_capacity, installation_analysis,
    optimal_padeye_analytical, cyclic_capacity_reduction,
)

anchor = SuctionAnchor(diameter_m=5.0, skirt_length_m=15.0,
                       padeye_depth_m=10.0,
                       submerged_weight_kN=250.0)
soil   = UndrainedClayProfile(su_mudline_kPa=5.0,
                              su_gradient_kPa_per_m=1.5)
load   = MooringLoad(tension_kN=4000.0,
                     angle_at_padeye_deg=25.0)

r = anchor_capacity(anchor, soil, method='dnv_rp_e303', load=load)
print(f"H_ult = {r.H_ult_kN:.0f} kN")
print(f"V_ult = {r.V_ult_kN:.0f} kN")
print(f"T_ult = {r.T_ult_kN:.0f} kN")
print(f"FoS   = {r.factor_of_safety(load.tension_kN):.2f}")

Capacity methods

Five independently testable methods behind a unified dispatcher op3.anchors.anchor_capacity():

method=

Reference

When to use

'dnv_rp_e303' (default)

DNV-RP-E303 (2021)

Code-compliant design baseline

'murff_hamilton'

Murff & Hamilton (1993)

Upper-bound cross-check

'api_rp_2sk'

API RP 2SK (2005)

Conservative API-style envelope (linear V-H)

'aubeny_2003'

Aubeny, Han & Murff (2003)

Detailed N:sub:`p`(z/D) for smooth/rough interface

'fe_calibrated'

OptumGX driver

Anchor-specific FE envelope (real CSV required)

Every method returns the same AnchorCapacityResult shape (H, V, T at the load angle + V-H envelope DataFrame + depth profile DataFrame). The FE method raises FileNotFoundError with a clear hint when its CSV is missing – the Python layer never fabricates FE data.

Installation analysis

installation_analysis() runs three checks per DNV-RP-E303 Sections 5.2-5.4:

  1. Self-weight penetration – bisection on \(W'_\text{sub} = R(z)\).

  2. Required suction vs depth: \(s_\text{req}(z) = (R(z) - W')\,/\,A_\text{lid,inner}\).

  3. Plug-heave stability ratio: \(R_\text{plug} = S_\text{pull}\,/\,S_\text{resist}\).

Cavitation limit

\(s_\text{allow}(z) = 0.9\,(\gamma_w\,z_w + \gamma'\,z)\) per RP-E303 Section 5.3.5 (0.9 cavitation margin for pump inefficiency + seawater vapour pressure).

Optimal padeye

Three approaches:

  • optimal_padeye_analytical(method='supachawarote_2005')() – Supachawarote et al. (2005) Table 2 interpolation by L/D and su profile shape.

  • optimal_padeye_analytical(method='murff_hamilton')() – constant 0.67 L per Murff & Hamilton (1993).

  • optimal_padeye_from_dissipation() – novel Op:sup:`3` contribution: centroid of the plastic-dissipation field at collapse, derived from OptumGX Mode D output. Extends Op3 Mode D (cavity-expansion framework) to anchor design.

The dissipation-centroid method requires a real dissipation.csv produced by op3/anchors/optumgx_anchor_run.py. See docs/ANCHOR_OPTUMGX_GUIDE.md for the LLM-assisted workflow.

Cyclic degradation

cyclic_capacity_reduction() applies the Andersen 2015 Drammen-clay surrogate

\[\frac{s_{u,\text{cyc}}}{s_{u,\text{static}}} = 1 - 0.45\,\bigl(\tau_\text{cyc}/s_u\bigr)^{1.2}\, \log_{10}(N) / \log_{10}(10^4)\]

calibrated to the four corner cases of Andersen (2015) FOG III Fig. 4 (N=10,1000 vs τ/s:sub:`u`=0.3,0.8). PI dependence is a small linear correction.

For a 3-hour storm with T:sub:`p`=12 s, τ/s:sub:`u`=0.5, this gives δ ≈ 0.86 (≈14% capacity drop).

MoorPy coupling

extract_anchor_loads_from_moorpy_system() solves MoorPy’s non-linear catenary equilibrium at every step of a prescribed body motion DataFrame (surge / sway / heave) and returns the real anchor-side tension + angle time series. anchor_safety_factor_timeseries() evaluates the capacity at each step and reports a per-step FoS against DNV-ST-0119 ULS = 1.30.

OptumGX driver

The commercial-boundary FE driver op3/anchors/optumgx_anchor_run.py is loaded into OPTUM GX’s desktop scripting console (it from OptumGX import * – not importable from a stock Python interpreter). It builds a 3D anchor model via 2D revolution, sweeps a configurable list of load angles (default 0/15/30/45/60/75/90 °), and writes:

  • envelope.csv – angle_deg, T_ult_kN_half, H_ult_kN, V_ult_kN

  • dissipation.csv – depth_m, w_z, D_total_kJ at the design angle

  • plates_a<ANG>.xlsx – raw plate-element data per probe

  • summary.json – run config + timings

The pure-Python post-processor op3.anchors.fe_postprocess.load_anchor_fe_results() then assembles AnchorCapacityResult + the dissipation-centroid padeye in one call.

Validation

  • Aubeny et al. (2003) Table 2 deep Np (smooth=9.14, rough=11.94)

  • API RP 2SK Section 5.4.2.3 cut-off (z/D=6 -> N:sub:`p`=9)

  • DNV-RP-E303 Section 4.3.3.2 Np profile

  • Closed-form \(H_\text{ult}\) integral with linear su

  • Randolph & House (2002) effective Np band for L/D ≥ 5

Total: 134 anchor tests passing (pytest tests/test_anchors/ -v).

API reference

op3.anchors.SuctionAnchor(diameter_m, ...[, ...])

Suction-anchor geometry and steel properties.

op3.anchors.UndrainedClayProfile(...[, ...])

Linearly increasing undrained shear strength profile.

op3.anchors.MooringLoad(tension_kN, ...)

Mooring-line load vector at the padeye.

op3.anchors.AnchorCapacityResult(method, ...)

Unified return type for every capacity method.

op3.anchors.anchor_capacity(anchor, soil[, ...])

Dispatch to the requested capacity method.

op3.anchors.installation_analysis(anchor, ...)

Run all three installation checks and return a single report.

op3.anchors.optimal_padeye_analytical(...[, ...])

Analytical / tabulated optimal padeye depth.

op3.anchors.optimal_padeye_from_dissipation(...)

Centroid of the plastic-dissipation field at collapse.

op3.anchors.cyclic_capacity_reduction(...[, ...])

Ultimate-capacity cyclic reduction factor for a 3-hour design storm.

op3.anchors.extract_anchor_loads_from_moorpy_system(...)

Solve MoorPy static equilibrium at every timestep of a motion history and return the anchor-side tension + angle.

op3.anchors.anchor_safety_factor_timeseries(...)

Compute factor of safety at every time step.

op3.anchors.generate_anchor_report(results, ...)

Render a Markdown design report from the FoS timeseries.

References

Andersen, K. H. (2015). “Cyclic soil parameters for offshore foundation design.” Frontiers in Offshore Geotechnics III, 5-82.

API (2005, reaff. 2015). RP 2SK: Design and Analysis of Stationkeeping Systems for Floating Structures, 3rd ed.

Aubeny, C. P., Han, S.-W., & Murff, J. D. (2003). “Inclined load capacity of suction caissons.” IJNAMG 27(14), 1235-1254.

DNV (2021). RP-E303: Geotechnical design and installation of suction anchors in clay.

Murff, J. D., & Hamilton, J. M. (1993). “P-Ultimate for undrained analysis of laterally loaded piles.” J. Geotech. Eng. 119(1), 91-107.

Randolph, M. F., & House, A. R. (2002). “Analysis of suction caisson capacity in clay.” OTC 14236.

Supachawarote, C., Randolph, M. F., & Gourvenec, S. (2005). “The effect of crack formation on the inclined pull-out capacity of suction caissons.” IACMAG Turin, Vol. 3, 577-584.