Along most of its length a buried pipeline carries a static load: the weight of the soil above it, plus internal pressure. Under a highway or a railroad it carries something different — a cyclic live load, applied and released with every axle, millions of times over the life of the line. Cyclic loading changes the failure question from "will it yield?" to "will it fatigue?", and it is why crossings get their own design standard, their own permit, and their own place on the critical path.
API RP 1102: the governing method
API RP 1102, Steel Pipelines Crossing Railroads and Highways, is the design procedure regulators and railroads expect to see. It uses a semi-empirical method — derived from full-scale instrumented crossing tests — that computes stresses in the carrier pipe from three sources and combines them:
- Earth load — the static circumferential stress from the weight of soil above the pipe, a function of cover depth, soil unit weight, and the pipe’s stiffness relative to the surrounding soil.
- Cyclic live load — the circumferential and longitudinal stress from surface wheel or axle loads, spread through the cover soil and amplified by an impact factor that accounts for dynamic effects.
- Internal pressure and thermal effects — the hoop and longitudinal stress the pipe already carries from operation.
The method then checks the combined effective (von Mises) stress against an allowable fraction of SMYS, and — critically for rail — checks the cyclic stress range at girth welds against a fatigue endurance limit. The API 1102 crossing calculator runs the full highway and railroad procedures for both cased and uncased configurations.
Why rail governs and highway usually does not
Railroad crossings are analyzed for Cooper E-80 loading (or whatever the railroad specifies), representing a standard heavy-axle train configuration. Highway crossings use design-vehicle wheel loads per AASHTO. The difference between them is substantial:
- Rail axle loads are heavier and more concentrated, and the crossties distribute them over a narrower footprint than a highway lane.
- Rail loading is more repetitive and more regular, which makes fatigue at girth welds a live concern rather than a theoretical one.
- Railroads impose their own engineering requirements on top of API 1102 — typically deeper cover, specified casing or an approved uncased design, defined workspace limits, flagging, and their own permit — and their review cycle is often the longest lead item in the crossing package.
The other structural driver is depth of cover, and it works in the direction that punishes shallow installations twice: as cover decreases, the live-load stress increases because the surface load has less soil to spread through, and the impact factor increases. This is exactly why crossings on the approaches to a bore or an HDD — where the profile is rising toward the exit and the cover is at its shallowest — are so often the governing case rather than the crossing itself. The depth of cover article covers how that profile interacts with the crossing check.
Cased or uncased?
Traditional practice installed the carrier pipe inside a larger steel casing at road and rail crossings. The intent was protective: the casing carried the live load, allowed the carrier to be replaced without disturbing the road or track, and contained a leak. Modern practice runs the other way, and most operators and regulators now prefer uncased crossings where the analysis supports them. The reasons are corrosion, not structure:
- A casing shields the carrier pipe from cathodic protection current, so the very section that is hardest to inspect or excavate is the section with the least corrosion protection.
- Casings fill with water and electrolyte over time, creating a corrosion cell around the carrier.
- Metallic short circuits between casing and carrier — from a shifted spacer or a settled end seal — are a chronic maintenance finding, and locating and clearing them is difficult and expensive.
- Casing end seals and vents deteriorate, and the annulus is not inspectable by in-line tools.
Against that, casing still makes sense where a railroad or highway authority requires it, where the crossing may need future replacement without disturbing the surface, or where installation method demands it — a bored or jacked casing installed first, with the carrier pulled through afterward, is a standard auger-boring configuration. Where casing is used, the design must include insulating spacers, sealed ends, vents, and a CP monitoring plan for the annulus. Where it is not, the uncased crossing analysis is the check that justifies leaving it out, and heavier wall through the crossing is the usual price.
Cover, permits, and installation method
Minimum cover at crossings is set by several overlapping authorities, and the deepest one wins:
- 49 CFR §192.327 sets the federal gas minimums: 30 inches of cover in normal soil in Class 1 locations, 36 inches in Class 2, 3, and 4 locations and at the drainage ditches of public roads and railroad crossings, reduced to 18 and 24 inches respectively where excavation is in consolidated rock.
- 49 CFR §195.248 does the same for hazardous liquids: 36 inches in industrial, commercial, and residential areas (30 in rock), 36 inches at drainage ditches of public roads and railroads, and 30 inches in any other area (18 in rock).
- Railroads typically require substantially more — commonly on the order of 5.5 feet from the base of rail to the top of the casing or carrier pipe, plus a separate minimum below any adjacent ditch, per AREMA practice and the individual railroad’s own standard.
- State DOTs and county road authorities set their own minimums within the highway right-of-way, usually deeper than the federal floor and measured from the ditch bottom rather than the pavement.
- Scour, future grading, and planned widening frequently govern over all of the above — a crossing designed to today’s ditch line can be left shallow by tomorrow’s road project.
Installation is almost always trenchless: open-cutting a highway or a rail line is rarely permitted. Auger boring with a casing is the standard for short crossings in stable soil, HDD for longer or deeper crossings and where the profile allows the entry and exit points to sit clear of the right-of-way, and jack-and-bore or direct pipe in between. The crossing permit from the road authority or railroad governs workspace, bore pit location, traffic control, monitoring, and often the allowable window of work. Those permits are typically the long-lead item on a crossing, and applying for them takes a design package that includes the API 1102 analysis.
The same problem, temporarily: construction traffic
One case regularly gets overlooked. During construction, heavy equipment crosses the newly installed line at designated points, often at less than final cover and with loads heavier than any highway vehicle the crossing was designed for — loaded rock trucks, sidebooms, tracked excavators, and crane counterweights. The analysis is the same wheel-load and track-load check, run against the as-built cover at that moment, and the mitigation is standard: designated crossing points, timber mats or steel plates to spread the load, temporary additional cover, and a rule that everyone crosses where they are told to. The wheel load and track load calculators size those temporary crossings, and the open-cut buried pipe load analysis covers the general earth-plus-live-load case for a trenched installation.