haulpave 0.4.0

Open-source pavement structure and operating-cost analysis toolkit for mine haul roads

HaulPave

Open-source pavement structure and operating-cost analysis toolkit for mine haul roads.

What HaulPave Does

• Computes traffic loading (CESA and design coverages) from a mining fleet composition.
• Determines pavement layer thicknesses using USACE CBR and TRH 14 empirical methods.
• Estimates comparative operating costs (tire, fuel, maintenance) across road surface scenarios.
• Generates structured design summaries with full versioning metadata.

What HaulPave Does NOT Do

HaulPave is not a complete haul road engineering tool. It does not cover geometric design, drainage, berm design, intersection layout, grade control, or traffic operations. All outputs require review by a qualified engineer. See SCOPE.md.

Installation

pip install haulpave

Quick Start

from haulpave.vehicle_registry import list_all, find_by_id
from haulpave.models.traffic import FleetUnit, TrafficInput
from haulpave.traffic.cesa import compute_cesa
from haulpave.pavement import cbr_thickness_from_coverages
from haulpave.economics import compare_scenarios, RoadScenario
from haulpave.reporting import build_design_summary

# 1. Pick a vehicle from the built-in registry
cat797f = find_by_id("cat-797f")  # Caterpillar 797F, 6104 kN GVW

# 2. Compute CESA from fleet
traffic = TrafficInput(
    fleet=[FleetUnit(vehicle=cat797f.vehicle, trips_per_day=30)],
    design_life_years=10,
    working_days_per_year=250,
)
result = compute_cesa(traffic)
print(f"Design CESA: {result.total_cesa:.2e}")

# 3. CBR pavement thickness
thickness_mm = cbr_thickness_from_coverages(
    subgrade_cbr=8.0,
    design_coverages=result.total_cesa,
)
print(f"Required thickness: {thickness_mm:.0f} mm")

# 4. Compare surface scenarios
scenarios = [
    RoadScenario(name="Asphalt", surface="asphalt", thickness_mm=thickness_mm,
                 haul_distance_km=10, trips_per_day=30),
    RoadScenario(name="Gravel",  surface="gravel",  thickness_mm=400,
                 haul_distance_km=10, trips_per_day=30),
]
comparison = compare_scenarios(scenarios)
for s in comparison.scenarios:
    print(f"{s.name}: fuel ${s.fuel_cost_usd_per_year:,.0f}/yr")

# 5. Build a versioned summary
summary = build_design_summary(
    inputs={"subgrade_cbr": 8.0},
    results={"thickness_mm": thickness_mm},
)
print(summary.package_version, summary.generated_at)

API Overview

Module	Key symbols
haulpave.vehicle_registry	list_all(), find_by_id(id), VehicleEntry
haulpave.traffic.cesa	compute_cesa(TrafficInput) → CesaResult
haulpave.traffic.coverages	compute_coverages(TrafficInput) → CoveragesResult
haulpave.pavement	cbr_thickness_from_coverages(), trh14_thickness_from_coverages(), design_pavement()
haulpave.economics	compare_scenarios([RoadScenario]) → ComparisonResult
haulpave.reporting	build_design_summary(inputs, results) → DesignSummary

Built-in Vehicle Registry

Four OEM mining trucks with GVW and 25/75 front/rear axle-load split (CAT Performance Handbook Ed. 47):

ID	Model	GVW (kN)
cat-797f	Caterpillar 797F	6 104
kom-960e	Komatsu 960E	5 925
cat-789d	Caterpillar 789D	3 304
cat-785d	Caterpillar 785D	2 641

Status

Phase	Content	Status
Phase 0 — Benchmark Foundation	Benchmarks 01–05, Pydantic models, USACE CBR curves	✅ Done (v0.1.0)
Phase 1 — MVP	CESA engine, coverages engine, CBR pavement design	✅ Done (v0.1.0)
Phase 2 — TRH 14 + comparison	TRH 14 engine, USACE vs TRH 14 comparison	✅ Done (v0.2.0)
Phase 3 — Economics + registry	Vehicle registry, rolling-resistance cost model, design summary	✅ Done (v0.3.0)
Phase 4 — Docs + community	MkDocs site, case studies, CLI polish	⏳ Planned (2026-Q4)

Design Principles

• Benchmark-first: every calculation method is tested against hand-computed examples before engine code is written.
• Confidence labeling: benchmark_tested / method_implemented / experimental on all outputs.
• Honest documentation: explicit assumptions, limitations, and method references.
• SI units internally: mm, km, kN, kPa, tonnes — no implicit unit mixing.
• Open-source hygiene: no proprietary data redistribution; OEM data requires source attribution.

License

MIT — see LICENSE.