Fabrication formulas

How to Calculate Weld Cost, Hydrotest Capacity, and NDE Hours for Oil and Gas Equipment

The core fabrication math for pressure vessels, spools, skids, and test bays, worked with real units, weld take-offs, and honest yield and uptime factors.

Pressure-boundary fabrication runs on five formulas, and every one starts with a take-off from a drawing, not a guess. Weld cost, diameter-inch spool pricing, skid assembly labor, hydrotest capacity, and NDE hours share the same skeleton: a physical quantity times a rate, adjusted by a scope or loss factor. Get the take-off and the rate right and the rest is arithmetic. This guide runs each calculation with real units and worked numbers so you can reproduce them from an isometric, a weld map, and a test schedule. Pricing and target ranges live in the sibling cost and benchmark guides; here the focus is purely on doing the math.

Start with pressure vessel welding. The Pressure Vessel Weld Cost formula is variable cost equals total weld length times cost per inch times weld scope, plus a fixed setup and NDE lump. Take weld length off the weld map in inches, summing longitudinal seams, circumferential seams, nozzle welds, and attachments. With 100 inches of weld at a blended 45 dollars per inch and 80 percent of the scope in your shop, variable cost is 100 times 45 times 0.80, or 3,600 dollars. Add a 250 dollar fixed setup and NDE charge and the total is 3,850 dollars, a blended 38.50 dollars per inch. The fixed lump stays constant because procedure qualification, fit-up, and PWHT do not scale with length.

Piping uses diameter-inches instead of linear inches so size and weld count collapse into one unit. A diameter-inch is nominal pipe size times weld count: one 6-inch butt weld is 6 DI, so ten of them is 60 DI. Pipe Spool Fabrication Cost multiplies total DI by cost per DI times scope, plus fixed material and handling. At 100 DI, 45 dollars per DI, 80 percent scope, and 250 dollars fixed, the variable is 3,600 dollars and the total is 3,850 dollars. Take DI straight off the isometric weld take-off. Carbon steel commonly sits near 30 to 60 dollars per DI, but stainless and exotic alloys need a separate rate band, not one blended number.

Skid labor is a rate problem. Skid Assembly Labor divides total components and connections by the crew assembly rate to get base time, then multiplies by an allowance factor. Count every discrete fit from the BOM and P&ID, not just major equipment: flange bolt-ups, instrument connections, valve installs, and supports. With 120 connections at 12 units per minute, base time is 120 divided by 12, or 10 minutes of pure fit time expressed as 10 hours of crew work once crew size and shift are baked into the rate. A 10 percent fit-up, staging, and rigging allowance lifts it to 11 hours. Congested or top-heavy skids need 15 to 25 percent.

Hydrotest capacity is where gross numbers lie. Gross capacity is units per cycle times available cycles; good capacity multiplies that by uptime and first-pass yield. Take units per cycle from the manifold position count and cycles from the shift hours divided by code hold time per cycle. With 4 units per cycle over 480 cycles, gross is 1,920. At 90 percent uptime you lose 192 units, and at 97 percent first-pass yield you lose about 52 more, leaving roughly 1,676 good units. Run Hydrotest Capacity and Valve Test Cycle Time together, since valve batches often share the same bay and each failed test burns a full retest cycle.

NDE hours follow the same rate structure but off the weld map and ITP. NDE Inspection Workload divides welds or joints to inspect by the inspection rate, then applies a setup, marking, and reporting allowance. The joint count is not the total weld count; it is the weld count times the required inspection percentage from your code or client spec, whether 10 percent spot RT or 100 percent RT. If a vessel has 200 welds at 100 percent RT, you inspect 200 joints; at 10 percent you inspect 20. Feed weld volume from the vessel and spool calculators straight into this one so inspection scope stays consistent with what you priced.

Valve testing mirrors skid labor. Valve Test Cycle Time divides valves to test by the test rate for base time, then applies a fixturing, changeover, and recording allowance. For 120 similar small valves at 12 per minute the base is 10 hours, and a 10 percent allowance gives 11. The trap is mixed batches: a large high-pressure gate valve with a long API 598 seat hold tests far slower than a small automated ball valve, so a single blended rate understates a mixed lot. Split the batch by size and class, rate each segment, and sum. Bake code-mandated hold times into the rate or the allowance rather than assuming back-to-back testing.

Two habits keep these calculations honest. First, tie every input to a source document: weld length and DI from the weld map and isometric, connection and joint counts from the BOM and ITP, uptime and yield from bay logs. Second, never let one rate cover unlike work. A carbon-steel DI rate on stainless, a fast fit rate on a congested skid, or 100 percent uptime on a bay with routine pump faults will each throw the answer off by double digits. When actuals diverge from the estimate by more than about 15 percent, the rate or the take-off is stale, and that is the number to rebuild before the next job.

Published 2026-07-01.