Oil and Gas Equipment Fabrication KPIs: Benchmarks for Weld, Test and NDE Performance

A pressure-equipment shop is judged on a handful of numbers: how much good work clears the test bay, how often welds pass inspection the first time, and how reliably it ships on the promised date. Use the ranges below as reference points for spotting a gap, not as targets, so you can tell whether a number is genuinely good or just familiar.

Every figure shows its basis. A superscript number links to a public source in the Sources list, and the sentence gives that source's scope. A figure marked no cited source is kept exactly as originally published because no public source stating it was found, so treat it as illustrative.

At a glance

First-pass hydrotest yield

First-pass hydrotest yield is the headline quality KPI because every failure burns a scarce bay cycle. World-class shops hold 98 to 99 percent first-pass yield on mature product no cited source; 95 to 97 percent is typical no cited source.

Anything below 95 percent signals weld defects, fitting leaks, or sealing problems worth a root-cause review no cited source.

The cheapest lever is upstream. A point of yield recovered in welding and assembly frees a full retest cycle in the bottleneck bay, which is worth far more than the same point measured at the bench.

How to measure

• Divide good tests by total tests attempted over the same scope.
• Separate first attempts from retests.
• Record the failure cause for every retest.

What moves it

• Weld quality upstream of the bay.
• Fitting and sealing practice at assembly.
• Root-cause review of every failed test.

See how yield and uptime set bay throughput with the Hydrotest Capacity calculator.

NDE reject rate

The NDE (nondestructive examination) reject rate is the leading indicator that hydrotest yield lags. The closest published reference is the weld repair rate: in TWI's 2011 industry survey, average repair rates for the oil and gas and power sectors ranged from 1 to 3 percent¹. For piping, pressure vessels and pipelines, the survey counts repaired welds over total welds.

A rate above 8 percent points to a procedure, position, or welder-qualification problem, not bad luck no cited source.

How to measure

• Count repaired welds over total welds examined.
• Track rejects per 100 welds by welder, position, and joint type so the pattern surfaces.
• Compare each welder and position with the shop's own average.

What moves it

• Retraining or requalifying the specific welders and positions driving rejects, rather than blanket inspection.
• Better weld access and fit-up, two of the main factors TWI names alongside welder skill¹.

Cutting the reject rate directly cuts rework hours and lifts downstream first-pass hydrotest yield within a job or two. Price the welding and NDE behind each vessel with the Pressure Vessel Weld Cost calculator.

Test bay uptime

Test bay uptime governs throughput even when quality is fine. A well-run hydrotest or valve bench holds 88 to 92 percent uptime no cited source; world-class disciplined maintenance and quick fixture changeover push 93 to 96 percent no cited source.

Below 85 percent, pump faults, fixture swaps, and unplanned maintenance are eating the schedule no cited source. Since the test bay is usually the true shipping bottleneck, a point of uptime here converts almost directly into shippable units.

How to measure

• Divide testing time by scheduled time, taken from bay logs.
• Log each stop as a pump fault, fixture swap or maintenance event.
• Report it per bay and per bench.

What moves it

• Preventive pump maintenance.
• Multi-position manifolds that raise units per cycle.
• Staged fixtures that cut changeover.

Model bench throughput with the Valve Test Cycle Time calculator.

Weld productivity

Weld productivity ties cost to schedule. The share of a welder's day spent with the arc on is the operator factor: arc-on hours divided by hours worked². Deposition efficiency is a different measure, the ratio of weld metal deposited to filler metal used².

Operator factor runs 20 to 35 percent world-class in a fab shop no cited source and 10 to 20 percent typical no cited source, with the gap lost to fit-up waits, material handling, and rework.

TWI's indicative duty cycles measure something narrower: arc burning time against the time a welder is working, with set-up and assembly excluded³. Read them as a ceiling on arc time, not as a whole-day benchmark.

Deposition rate depends on process. FCAW fill on heavy wall runs 8 to 15 pounds per hour no cited source, inside TWI's indicative FCAW range of 1.0 to 15 kg/h, about 2.2 to 33 lb/h, measured welding continuously without breaks⁴.

Cold-wire GTAW deposits much less. A 2023 peer-reviewed review reports that at 6000 W of arc energy, the maximum deposition rate for cold-wire GTAW on steel is about 2 kg/h, roughly 4.4 lb/h⁵.

How to measure

• Divide arc-on hours by hours worked, per welder and per cell.
• Log the waits that keep the arc off, such as fit-up and material handling.
• Track deposited weight per arc-on hour by process.

What moves it

• Fit-up sequencing, usually the biggest loss.
• Part staging at the cell.
• Less rework, which frees arc time for new joints.

Attack the biggest loss before chasing a faster process; higher arc-on time shortens the weld critical path that feeds both NDE and testing. Cost a spool's welds by diameter-inch with the Pipe Spool Fabrication Cost calculator.

On-time delivery

Schedule KPIs decide whether the shop is trusted with the next bid. On-time delivery of 95 percent or better is world-class no cited source; 85 to 92 percent is typical no cited source, and repeated slips trace almost always to long-lead castings, not welding.

The dominant lever is procurement discipline: order castings and forgings against their real 12 to 30 week lead no cited source with a service-level buffer, and hold NDE and hydrotest float in the plan. A shop that protects the casting date and the test bay hits its dates; one that reacts to slips never catches up.

How to measure

• Measure delivery against the original committed date, not the last revised one.
• Count late orders by root cause, such as castings, NDE or test bay.
• Review it weekly alongside open purchase orders for long-lead parts.

What moves it

• Ordering castings and forgings against their real lead time.
• A service-level buffer on long-lead items.
• Planned float for NDE and hydrotest.

Check assembly hours against the committed date with the Skid Assembly Labor calculator.

Field service and rental fleet KPIs

Aftermarket and rental KPIs carry the margin between projects. Field service margin, the share of service jobs hitting their margin target, should sit at 90 percent or better world-class no cited source and 75 to 85 percent typical no cited source; a persistent gap means travel, mobilization, or scope creep is eating billable time.

On rental fleets, refurbishment cost per unit and the share of the fleet actually needing work drive the budget. World-class operations keep 10 to 20 percent of the fleet in refurb per cycle no cited source, while a poorly maintained fleet pushes 30 to 50 percent no cited source.

How to measure

• Divide service jobs that met their margin target by all service jobs closed.
• Count fleet units sent to refurbishment in each rental cycle.
• Track refurbishment cost per unit alongside that share.

What moves it

• Tighter turnaround scoping before the job starts.
• Condition-based, not calendar-based, recertification.

How to build your own baseline

Improvement follows a fixed priority: fix the bottleneck KPI first, then the one feeding it. In this category that usually means the following order.

• Protect test bay uptime and first-pass hydrotest yield.
• Drive the NDE reject rate down upstream.
• Recover welder arc-on time to shorten the weld path.
• Lock procurement to defend on-time delivery.
• Review all of them on a tier board weekly, with the same denominators every time.

For example, a 96 percent yield this week is comparable to last week's only if both weeks count tests over the same scope.

A shop that holds 98 percent first-pass yield no cited source, 93 percent bay uptime no cited source and 95 percent OTD no cited source is genuinely world-class, and each of those numbers is a target you can measure against tomorrow. For weld rejects, the cited reference point is the 1 to 3 percent average repair rate TWI's 2011 survey found for the oil and gas and power sectors¹.