Process reasoning
Define the batch or production period represented by each sample. For a continuous plant, collect timed increments while recording feed rate and operating conditions. Keep unusual feed states separate when mixing would hide an important limitation. Settling, cooling, evaporation and chemical ageing can change a slurry between collection and testing; record those changes rather than treating the delivered sample as an exact copy of the process. A composite is useful for an average composition but cannot prove performance during a short disturbance.
ILLUSTRATIVE EXAMPLE
Apply the reasoning
Suppose a plant normally supplies 8 wt% solids but receives a 14 wt% solids flush for twenty minutes each shift. A single blended sample may produce an apparently manageable average. Test the normal feed and the flush separately, then calculate whether equalization or a restricted feed rate is needed. State the duration and frequency of the flush in the equipment design basis. These values illustrate a sampling decision, not an allowable HUADA feed range.
What to record
| Item | Basis to document |
|---|---|
| Sample identity | Batch, time, process location and collection method |
| State at collection | Temperature, solids concentration, density and visible phase separation |
| Handling history | Storage duration, mixing, dilution and any temperature change |
| Required outputs | Cake moisture basis, suspended-solids limit, purity and recovery |
Review checklist
- Agree sample quantities and shipping arrangements with the test facility before dispatch.
- Record normal and difficult feed states separately.
- Reserve material for repeat measurements and feed characterization.
- Use consistent analytical methods for feed and all outlet streams.
Your next decision
Ask for a report that links every result to a sample ID and a documented operating condition. Repeat the reference condition after the trial sequence to detect a feed change or progressive fouling. If the intended worst-case feed was not tested, keep that gap visible in the recommendation instead of turning the best result into a guarantee.