Module 6.6 — Isokinetic sampling in depth, building on Module 1.4's introduction, applied specifically to the unique challenge of sampling steam.
Click either sample to see how velocity mismatch biases the reading. A too-slow sample pulls in a disproportionate share of heavier moisture droplets.
Module 1.4 introduced isokinetic sampling briefly, noting it "matters most for steam sampling, where a mismatched sample velocity can bias particulate and moisture carryover readings." Everything covered since then — carryover mechanisms (6.2), silica volatility (6.3), turbine deposition (6.4) — depends on trustworthy steam purity data to detect and diagnose. This module explains why steam sampling specifically demands more sampling rigor than liquid sampling, and what happens when that rigor is missing.
Liquid sampling (covered generally in Module 1.4) draws a relatively homogeneous single-phase fluid — any reasonable sample point and flow rate gives a representative reading. Steam, particularly steam containing any moisture (recall the Wilson line concept from Module 6.4), is fundamentally different: it's a two-phase flow, with vapor and liquid droplets moving through the pipe together but not necessarily at identical velocities or in identical concentration across the pipe cross-section. A sample drawn carelessly from this two-phase flow doesn't necessarily represent the true average composition of what's actually flowing through the pipe.
Liquid moisture droplets carry more inertia than the surrounding vapor due to their higher density. If a sample nozzle draws steam at a velocity slower than the bulk process flow, droplets — following their inertia rather than smoothly following the streamlines that bend into the slower-moving nozzle — deflect into the sample at a higher rate than their true proportion in the bulk flow, artificially inflating apparent moisture and carryover readings. If the sample is drawn faster than process velocity, the opposite bias occurs: droplets are underrepresented, and the sample reads artificially cleaner than actual conditions. Isokinetic sampling — matching sample extraction velocity precisely to local process velocity — is the only way to avoid both biases.
Beyond velocity matching, sample nozzle position within the steam pipe cross-section matters, since two-phase flow isn't necessarily uniform across the pipe — moisture concentration can vary between the pipe center and the walls, particularly after a bend or other flow disruption. Standard practice positions steam sample nozzles at the pipe centerline, in a straight run of pipe well away from bends or other disruptions, specifically to draw from a location most representative of the overall flow rather than a potentially biased near-wall region.