Module 1.4 — Grab samples vs. online instrumentation, sample panel basics, and why sample point selection is a diagnostic decision, not just a checklist item.
Click any stage to see its role in getting a trustworthy reading. Every sample must be cooled to a standard reference temperature before any reading is trusted.
Every number covered in Modules 1.1–1.3 is only as trustworthy as the sample it came from. A perfectly executed treatment program can look like it's failing if the sample delivery system is fouled, uncooled, or pulling from the wrong point — and a real excursion can hide behind a bad sample just as easily. Understanding how samples get from the process to a reading is not optional background; it's part of reading the chemistry correctly at all.
Online (continuous) instruments sit in a sample panel and feed real-time readings to the plant's data system — pH, conductivity, dissolved oxygen, and sodium are the parameters most commonly monitored this way because trends matter as much as single readings, and a slow drift is often the earliest warning of a problem. Grab samples are manually collected and either read on portable equipment or sent to the lab, used for parameters that are harder to automate reliably (certain silica test methods, some hardness tests) or to verify an online instrument's reading when it's in question.
Conductivity and pH are both temperature-dependent measurements. Industry-standard practice is cooling samples to a fixed reference temperature (commonly 77°F / 25°C) before they reach an instrument, using a sample cooler. Skip or degrade this step — a fouled cooler, insufficient cooling water flow — and every downstream reading shifts, sometimes enough to mask a real excursion or manufacture a false one. A chemistry excursion investigation that doesn't start by confirming the sample cooling system is working correctly is starting in the wrong place.
Where a sample is drawn from matters as much as how it's measured. Recall the sample points from Module 1.1's diagram — deaerator outlet, economizer inlet, steam drum: each exists because it isolates a different part of the cycle and a different failure mode. A sample drawn from the wrong point, or a sample line that's too long or poorly insulated (letting the sample degas or change temperature before reaching the panel), can give a technically accurate reading of the wrong thing.
Online instrument readings typically feed into the plant's distributed control system (DCS) and get archived in a data historian, which is what makes trending possible — spotting a slow six-hour drift in cation conductivity that a single grab sample would never catch. Knowing how to pull and read a trend, not just a current value, is often the difference between catching an excursion early and catching it after it's already caused damage.