Assumptions, limitations and open questions¶
BoreTHM screens a stated drilling programme using the supplied inputs. A failed assessment does not establish that no other well design or existing technology can drill the site. The validation guide describes what each comparison and check establishes.
Input evidence¶
The deep targets extend beyond the measured stress and temperature records. Stress extrapolation and the strength assumptions affect the pressure requirements. At United Downs, strength is calibrated using an assumed stress profile, so that comparison cannot independently verify both quantities. Larderello and Pannonian retain unsourced strength assumptions. The site-input records identify the sources and uncertainties.
The circulation comparisons use FORGE observations at depths and temperatures below the deep targets. Newberry's well check reproduces the volcanic breakouts but also predicts failure in granodiorite where none was logged. These checks do not establish the accuracy of a complete deep drilling programme.
Applicability and numerical guards¶
Input acceptance, property-table coverage and field evidence impose different limits:
| Item | Range or treatment |
|---|---|
| Accepted site inputs | validate_site applies the scalar ranges below to catch unit slips and implausible inputs. It checks a finite geotherm sampled every 100 m, including the target; the sampled open hole must warm with depth, and the target temperature must agree within 5 °C. It samples stress profiles every 500 m from the target to the casing shoe, including the shoe, checking finite positive stresses, maximum horizontal stress at least equal to minimum horizontal stress, the ordering of any declared faulting regime, and pore pressure below the minimum horizontal stress. It also checks provenance labels and reports missing sources. These checks do not establish that every fluid state remains inside the property table. |
| Water-property table | water_properties, implemented by water_table.props, interpolates over 1 to 479 °C and 0.1 to 218.1 MPa absolute. Each grid has a spacing of 2 in its stated units. Inputs outside those limits are clipped to the nearest edge before interpolation. These are the cached grid's limits. |
| Other numerical guards | The water-only heat-capacity path, model1_coupled._cp, clips its inputs to 1 to 480 °C and 0.1 to 220 MPa absolute, then calls the heat-capacity interpolator directly. That interpolator extrapolates over the small intervals beyond the final grid nodes at 479 °C and 218.1 MPa. model1_coupled.solve marks a circulation solution unsuccessful if either sampled fluid profile reaches 480 °C. The friction-heating lookup in model1_coupled._friction_gradient uses a minimum property pressure of 5 MPa absolute and clips temperature to 1 to 480 °C. The separate hydraulics function uses the approximate hydrostatic property pressure directly. |
| Fluid model | Circulation represents one single-phase fluid. The weighted-fluid model treats barite as a dispersed solid in water. There is no two-phase liquid-vapour flow calculation or vapour-fraction balance. |
| Circulation property pressure | model1_coupled.P_of_z supplies the approximate absolute pressure P_surface + 10 MPa/km × TVD, with true vertical depth (TVD) measured from surface and a default surface pressure of 0.1 MPa absolute. This property pressure is separate from the wall-pressure budget. The property-pressure comparison, A1, tests a different treatment at fixed programme inputs. |
| Field comparison range | The FORGE 16B circulation trial spans about 2.31 to 2.99 km measured depth, or 2.12 to 2.41 km true vertical depth. Its adopted formation-temperature profile, from neighbouring well 16A, gives about 182 to 205 °C over that interval. These conditions are below the deep reference targets; the validation guide describes the measured fluid temperatures and fitted parameters. |
| Site conditions | Deep stress and temperature profiles extend beyond the measured intervals. Strength assumptions and the transfer of calibrated parameters remain site-specific uncertainties, as described in input evidence. |
Accepted scalar ranges¶
The inclusive limits below are those in site_evaluation.INPUT_RANGES. Unconfined compressive strength (UCS) is supplied at 25 °C. Invalid values raise ValueError; missing provenance produces warnings.
| Input | Accepted range | Unit |
|---|---|---|
| Effective target depth | 500 to 20,000 | m |
| Target rock temperature | 50 to 600 | °C |
| Rock conductivity | 0.5 to 7 | W/(m K) |
| Rock Young's modulus | 5 to 150 | GPa |
| UCS and every strength-case UCS | 10 to 500 | MPa |
| Injection temperature | 0 to 100 | °C |
| Friction coefficient | 0.2 to 1.5 | dimensionless |
| Reference fluid-pressure gradient | 8,000 to 20,000 | Pa/m |
| Reference vertical-stress gradient | 15,000 to 35,000 | Pa/m |
These scalar limits are distinct from the temperature and pressure domain of the property table.
Breakout formation and cooling¶
The circulation rock term uses a minimum exposure of one hour for a depth-dependent history, or 0.001 years (about 8.8 hours) for a uniform t_years. This is a restriction of its quasi-steady line-source approximation. The separate radial wall-temperature calculation resolves early cooling with its supplied state durations, using fluid-temperature and film-coefficient boundaries from circulation. It does not apply the circulation exposure floors. Those floors set no breakout formation delay. The circulation equations describe their scope.
The base case assumes breakouts form as soon as the bit exposes the rock. A skin-depth heuristic approximates where failure begins behind the wall; immediate formation prevents cooling from reducing the initial drilling pressure requirement. The physical-model guide explains that approximation.
The breakout-timing study tests assumed delays between exposure and failure. Its results show when the whole-hole drilling-fluid window opens under those assumptions. They do not measure breakout formation time. Images showing breakouts within minutes in sediment do not establish timing or growth in hot crystalline rock. Near-bit pre-cooling and the growth of breakouts remain open questions.
Fluid properties and equipment qualification¶
The fluid model represents one reference mud. Its yield point is fixed and its plastic viscosity is scaled from water. Barite heat capacity and thermal expansion are treated as constants. The design study's barite source check gives the supporting values and their limits.
A published potassium-formate fluid retained a plastic viscosity of 30 mPa s after ageing at 200 °C, measured after cooling to 49 °C. That result does not qualify the modelled mud during a 400 °C trip. The fluid and equipment sources describe the conditions behind the comparisons.
The manufacturer's 204 °C coating rating does not establish survival through a hotter drilling cycle. The cited managed-pressure drilling equipment has a stated mechanical pressure rating; qualification for the modelled thermal cycle remains unknown.
Tool temperature and programme timing¶
The tool is represented by a lumped thermal model of the collar and enclosed fluid. During a staging pause, coolant is supplied at its steady temperature without accounting for its travel time through the string. Staging duration is absent from the base trip clock. During run-in, the trip-fluid temperature column is frozen and a uniform heavier fluid is assumed present from the start.
These approximations matter when interpreting a small tool-temperature margin or a narrow allowable density range. The approximation comparisons test selected alternatives. Their results do not replace a calculation with a common timeline and explicit fluid displacement.
Other design limits and study scope¶
The assessments do not include detailed casing and cement integrity, string loads and fatigue, circulation losses, or creep coupled to the drilling cycle. The hole-cleaning reference and cycle timings come from one well. Rig capacity is checked against three pumps of one published rating. The 0.02 SG allowance for pressure changes during pipe movement is a sensitivity assumption, with no verified field range for these 10 to 13 km wells.
The approximation study generally changes one assumption at a time and does not cover every joint uncertainty. The design study samples selected settings and one pair per site. Neither study makes either deep-site programme pass or establishes a complete operating envelope.
The conditions for further implementation work are recorded in the maintenance priorities. The study summary lists questions for specialist discussion.