Define the feed condition and the turndown case
A falling-film trial often begins with a target outlet concentration and a feed rate. Those numbers are necessary, but they do not describe what the distributor has to do. The trial should start with a written feed condition: the solids or Brix basis and its measurement method, feed temperature, viscosity at the relevant concentration and temperature, suspended particles and fibres, foaming tendency, acidity, and the range of compositions expected during the campaign.
Fruit processing adds variation before the evaporator. The FAO guidance on small- and medium-scale juice processing describes fruit-specific preparation, from washing and sorting to extraction, screening, deaeration and clarification, and notes that cultivar, maturity and pre-treatment change the material reaching later stages. A trial organised around one clear juice sample may therefore say little about a cloudy, pulpy or partly fermented feed. Read the FAO juice-manufacture chapter.
The Tetra Pak Dairy Processing Handbook notes that product properties such as viscosity and heat stability determine how far concentration can be forced. That is a principle, not a fruit-specific setting. Ask for the properties at the most concentrated case and at the lowest planned load. Read the evaporator chapter.
State turndown as a real operating case: lowest planned rate, start-up and shutdown sequence, product changes and the duration of each condition. Do not borrow a minimum-load percentage from another product. The trial question is whether the distributor, heating surface and separator remain workable at the edges of this duty.
Inspect what the distributor is being asked to do
The handbook describes uniform distribution over the heating surface as important for economical operation. Gaps in distribution can cause local overheating, product sticking, impaired heat transfer and more difficult cleaning, reducing production uptime. Its tubular example uses a divider and a drilled distribution plate above the tubes. Its plate example uses spray nozzles and introduces product at evaporation temperature to avoid flash evaporation during distribution.
Those details belong in the trial because the distributor is being asked to work across a range, not at one decorative flow. Record the proposed hole or passage geometry, the expected pressure at the head, the feed temperature relative to boiling, and the way the head is opened for inspection. Ask how the design handles particles, fibres, foam and any flashing at the inlet.
A total feed-flow reading cannot prove that every passage is wetted. The protocol should state which checks are practical: distributor inspection before and after the run, flow verification by an agreed method, and observation of pressure and temperature responses when load changes. If the design cannot be inspected while running, plan a controlled shutdown check rather than implying continuous verification.
Compare the product path with the vapour path
The product and vapour paths separate at the heating-section outlet. Vapour may heat another effect, enter a thermocompressor or mechanical compressor, or go to a condenser. Each route has a different consequence for pressure, carry-over and the evidence needed at start-up.
The handbook notes that as evaporation proceeds the liquid volume falls and the vapour volume rises. If vapour velocity exceeds the available space, pressure drop increases and more temperature difference is needed. Separator and duct design therefore belong to the trial, especially at low pressure and high vapour load.
Carry-over is not only a product-loss question. Droplets can contaminate recovered condensate or reach compression equipment. The trial should identify sample points for feed, concentrate and condensate, and record separator level, pressure and non-condensable extraction alongside them. A clear-looking condensate is not an analytical result.
Draw the product and vapour routes separately in the protocol, marking sample points, vents, drains and recycle streams. That reduces misattribution and gives the cleaning team a boundary to verify. It also shows where a future aroma-recovery or deaeration interface would sit, without implying those optional duties are included.
Record start-up and cleaning evidence, not only steady state
Steady-state data are useful but incomplete. Start-up exposes the sequence that steady operation hides: evacuation, preheating, first feed, establishment of the film, ramp to concentration, a low-load or product-change case, shutdown, drain and cleaning.
Record time-stamped feed flow, temperatures, pressures, levels and concentration at each planned step. Note deviations and who responded to them. If the trial includes a turndown case, define the lowest load and how long it will be held. Do not assume that a stable reading at normal load predicts the low-load case.
Cleaning needs its own evidence. The handbook lists cleaning duration and cost among the planning aspects of an evaporator station. The FAO guidance stresses equipment and water sanitation and the risk of contamination during preparation. For the trial, agree the cleaning chemistry, temperature, contact time and return path, then inspect the distributor, heating surface, separator and transfer line afterwards.
Record what changed after cleaning: inspection findings, rinse acceptance, next start-up performance and any repeat deposit location. Those observations are more useful than a general easy-to-clean claim.
Agree what the trial is allowed to prove
A trial can establish how one representative feed behaved under a defined sequence with named measurements. It cannot guarantee year-round composition, shelf life, nutrient retention or a universal minimum wetting condition. The FAO guidance warns that carefully controlled laboratory or pilot work does not automatically resolve the logistics and variability of commercial scale.
Dairy examples in the handbook are not fruit settings. Its pressure and temperature examples, effect arrangements and utility figures come from dairy processing; they illustrate principles and equipment relationships, not a specification for juice, extract or another food liquid.
Write the acceptance basis before the trial. It should name the feed, operating cases, measured streams, utility conditions, product checks and cleaning verification. Where a value is unknown, record it as a gap. A report that presents a single outlet concentration without those conditions is not a sufficient basis for a performance guarantee.
Questions to put in the trial protocol
- Which feed properties are measured, at what temperature and concentration, and which are still unknown?
- What is the lowest planned load, and how will distribution be checked there?
- Where are the product, vapour and condensate sample points?
- Which separator and carry-over measurements are recorded, and what action follows an adverse result?
- What start-up, shutdown and cleaning steps are included, and who witnesses them?
- Which results are acceptance criteria, and which are only observations?
Bring the proposed protocol to the project review before the equipment scope is fixed. The distributor, separator, vacuum package and cleaning boundary should be discussed as one connected duty. Discuss a falling-film trial