Process Skid Fabrication KPIs: Weld Reject, FAT Yield and Delivery Benchmarks

Use these benchmarks as reference points for spotting a gap worth investigating, not as targets. Skid work varies from one design to the next, so your own trend across recent builds says more about your shop than any single published figure.

Every figure shows its basis. A superscript number links to a public source in the Sources list, and the sentence gives that source's year and scope. A figure marked no cited source is kept exactly as originally published because it has not been matched to a public source that states it, so treat it as illustrative.

At a glance

Weld reject rate

Weld reject rate is the KPI that governs schedule and quality on a skid shop floor. Measure it as rejected joints divided by joints inspected, tracked by welder and by process.

Published survey data gives a reference point. In a 2011 industry survey reported by TWI (The Welding Institute), the average repair rate for piping systems was 3.0 percent of welds¹. The same source puts average repair rates for the oil and gas and power sectors at 1 to 3 percent, with peaks up to 25 percent in specific locations¹.

World-class carbon-steel cells running gas tungsten arc (GTAW) plus shielded metal arc (SMAW) welding hold 1 to 2 percent reject no cited source. Anything above 6 percent signals a welding procedure specification (WPS), fit-up or welder-qualification problem worth stopping the line for no cited source.

Each repair costs two to three times the original weld no cited source, so the rate drives cost as well as schedule. Track it live off the nondestructive examination (NDE) log, not at project close.

For example, cutting reject from 5 to 2 percent on a 600 diameter-inch skid means 3 fewer repaired joints in every 100 inspected, which recovers days of critical-path time.

How to measure

• Divide rejected joints by joints inspected.
• Split the rate by welder and by welding process.
• Update it from the NDE log as results arrive, not at project close.

What moves it

• Control of the welding procedure specification.
• Fit-up quality before welding starts.
• Welder qualification and targeted retraining.

Size the weld workload behind the rate with the Weld Inch Estimate calculator.

Bay utilization

Bay utilization tells you whether floor space is turning work or holding it. The KPI is value-add time divided by staffed bay hours, and each recovered point adds real skids without new floor space.

Well-run modular shops sustain 80 to 90 percent no cited source. Below 70 percent points to crane bottlenecks, material staging delays or rework loops parked in the bay no cited source.

How to measure

• Divide value-add time by staffed bay hours.
• Take the losses from downtime logs, not estimates.
• Separate crane wait, inspection holds and material starvation into their own buckets.

What moves it

• Attacking the largest loss bucket first rather than chasing the averaged number.
• Crane availability when the bay needs a lift.
• Material staged before the crew needs it.
• Keeping rework loops out of production bays.

Frame the throughput view of each bay with the Skid Footprint calculator.

First-pass FAT yield

First-pass FAT yield measures how many skids pass the witnessed factory acceptance test (FAT) with no major hold. A mature repeat design should clear 85 to 95 percent first-pass no cited source; first-of-a-kind skids realistically land at 60 to 75 percent no cited source.

Pair it with a count of open punch items per skid. World-class shops carry under 5 open items into a witnessed FAT no cited source, while a struggling line walks in with 20-plus and burns the schedule live no cited source.

The lever is pre-FAT dry runs and closing punch items before the client arrives. Padding the witness window only hides the problem.

How to measure

• Divide skids passing clean by skids tested.
• Count open punch items per skid on the day the witness arrives.
• Report repeat designs and first-of-a-kind skids separately.

What moves it

• Pre-FAT dry runs in the shop.
• Closing punch items before the client arrives.

Plan the witness window with the FAT Duration calculator.

Schedule adherence and on-time delivery

Schedule adherence, measured as skids shipped on the committed date divided by skids due, is the KPI clients actually feel. A 75 to 85 percent on-time rate is typical no cited source, and chronic sub-70 percent usually traces to a single gating stage, most often piping fabrication or FAT no cited source.

For a cross-industry reference, the IndustryWeek Census of Manufacturers, a spring 2000 survey of more than 3,300 U.S. plants, defined delivery leaders as plants with on-time delivery of at least 98 percent². Only about one-third of the participating plants reached that level².

Diagnose a miss by comparing the Pipe Spool Count net deliverable rate against the cut-list requirement. If net capacity trails the schedule, no amount of expediting downstream recovers it.

How to measure

• Divide skids shipped on the committed date by skids due.
• Measure against the committed date, not a revised one.
• Tag each miss with the stage that gated it.

What moves it

• Weld station uptime, upstream of everything else.
• First-pass yield in welding and at FAT.
• Piping fabrication capacity measured against the cut list.

Check net spool capacity against the schedule with the Pipe Spool Count calculator.

Instrument loop-check productivity

Instrument and control (I&C) productivity separates commissioning cost leaders from laggards. Track loop-check labor per loop and terminations per technician-hour.

A well-run shop checks simple analog loops in 1.5 to 3 hours and lands per-loop labor near 30 to 60 dollars no cited source. A poorly staged one drifts toward 80 to 100 dollars as fixed mobilization spreads across too few loops per trip no cited source.

Benchmark per-loop cost across your last several skids. A rising trend flags either creeping loop complexity or a fixed cost getting thin, and both are fixable before the next bid.

How to measure

• Divide loop-check labor hours and cost by loops checked.
• Count terminations per technician-hour.
• Compare per-loop cost across your last several skids.

What moves it

• Pre-calibrating instruments in the shop.
• Batching loops per site visit so mobilization spreads across more loops.

Estimate loop labor per skid with the Instrument Loop Labor calculator.

Rework hours

Rework and contingency burn is the health check on your estimating discipline. The KPI is rework hours actually spent divided by total fab hours, read against the correction hours you budgeted.

World-class skid builds hold rework under 3 percent of fab hours no cited source. A share of 5 to 8 percent is common no cited source, and double digits means quality is being inspected in rather than built in no cited source.

A shop that closes the loop from NDE reject data back to welder training and WPS control drives this number down build over build. One that does not keeps rediscovering the same defects.

How to measure

• Divide rework hours spent by total fab hours.
• Trend it by skid and by root cause.
• Compare consumed correction hours against the rework allowance in the estimate.

What moves it

• Feeding NDE reject data back into welder training.
• Welding procedure control.

Estimate the welding scope your fab hours rest on with the Weld Inch Estimate calculator.

Dispatch readiness

Shipping and dispatch cadence gates the whole modular economic case, because a plant schedule is paced by module flow, not fabrication speed. Track transport-ready availability: the share of a load-out window when crane, riggers and permits are actually ready.

Well-run yards with staged permits and dedicated rigging run 88 to 95 percent no cited source. Below 80 percent you are losing more than a shift a week to crane wait or oversize-load holds no cited source.

Modules stacking in the laydown yard tie up capital and crane time. The lever is staging permits ahead of the pick and pre-building securement kits, not adding cranes.

How to measure

• Divide the load-out window hours when crane, riggers and permits are all ready by total window hours.
• Log each hold by cause: crane wait, rigging or permits.

What moves it

• Staging permits ahead of the pick.
• Pre-building securement kits.
• Dedicated rigging crews for load-out.

Check each module's load envelope with the Module Shipping Cube calculator.

Improve them in the order they gate throughput

Weld reject and bay utilization come first, because they sit on the fab critical path. First-pass FAT yield and schedule adherence, which the client sees, come next, and then I&C productivity and dispatch cadence, which govern delivered cost.

Set a single owner per metric, review the trend weekly against the world-class band, and tie each metric to a concrete lever with a number attached.

For example, a shop that moves reject from 5 to 2 percent, utilization from 72 to 85 percent and first-pass FAT from 70 to 88 percent will out-ship and out-margin a competitor working from the same drawings.

How to build your own baseline

• Write down each KPI's definition and keep it fixed.
• Split weld reject by welder and process, and FAT yield by repeat versus first-of-a-kind design.
• Collect data across several skids before you set a target.
• Compare against your own trend first and published ranges second.