Spray Rate Control

Spray rate is the mass (or volume) of solution delivered per unit time, typically expressed per nozzle. A four-nozzle fluid-bed running at 80 g/min per nozzle is delivering 320 g/min total binder solution to the bed. The rate has to be matched to the bed's evaporative capacity — set by inlet air temperature, inlet air flow and inlet humidity — so that the bed neither over-wets (collapse, agglomeration, sticking to vessel walls and probes) nor under-wets (no granulation, just spray-drying of binder droplets in mid-air that then coat onto particles as dry binder, never wetting the core).

Two regimes describe the operating envelope. In the constant-rate regime — spray rate ≤ evaporative capacity — the bed is at a stable wet equilibrium and granule growth proceeds steadily. Above the evaporative capacity, the wet equilibrium tips: the bed temperature drops, agglomerates grow uncontrollably, sticking to internal surfaces begins, and within minutes the process can collapse to a static lump that requires manual recovery. Below ~50 % of the evaporative capacity, spray-drying dominates — binder droplets dry to a fine fines tail in the bed rather than depositing on particle surfaces — and the granulation looks like a coating step that produced no granules.

• Closed-loop control via peristaltic pump (the dominant choice for ease of CIP and tubing change) or piston pump (for higher pressure or higher accuracy).
• Per-nozzle metering — single-flow-meter multi-nozzle setups hide imbalance.
• Pump calibration before each campaign — peristaltic tubing wears unpredictably and rate drifts.
• Spray rate profile (ramp-up at start, hold during constant-rate, ramp-down approaching endpoint) is part of the control recipe, not a single setpoint.
• Redundant cross-check via load-cell mass loss on the solution tank — catches flow-meter failure that would otherwise be silent.

These ranges are descriptive of the operating envelope; the validated spray rate for a specific product is always a narrow band within the broader range, established in PPQ and locked into the recipe. A common error is to read the broad range as 'somewhere in here is fine' and skip the validation work that pinpoints the actual operating window for the product.

Spray rate sets one input. The bed's response depends equally on atomisation air pressure (which sets droplet size — typically 20 – 100 µm Sauter mean diameter), inlet air temperature (which sets droplet evaporation in flight), and bed surface refresh rate (which sets the residence time of each particle near the spray zone). Three of these have to align.

Droplet size relative to particle size determines the granulation mechanism. Droplets much smaller than particles (e.g., 30 µm droplets onto 200 µm particles) coat by 'distributed deposition' — many droplets per particle, building a thin layer; this is the layering or coating regime. Droplets comparable to particles ('immersion' regime) wet whole particles to saturation — this is the granulation regime, where wet bridges form between particles and growth is nucleated. Spray rate and atomisation pressure together set which regime dominates: high spray rate + low atomisation pressure = large droplets = immersion (granulation); low spray rate + high atomisation pressure = fine droplets = distributed deposition (coating).

Inlet air conditions act as a moderator. Hot, dry inlet air evaporates droplets in flight before they reach the bed — a real risk above 80 °C inlet with aqueous solution at low spray rate. Cool, humid inlet air leaves droplets oversized and wet on arrival, increasing the wet-equilibrium risk. The recipe locks the four-way balance: spray rate × atomisation pressure × inlet temperature × inlet humidity (when available). Each is a CPP in its own right.

1. Calibrate the peristaltic pump (or piston pump) before each campaign — measured against a tared scale over a defined volume; record the result in the batch.
2. Inspect every nozzle before phase start; a blocked or partially blocked nozzle destroys per-nozzle balance and creates local over-spray.
3. Confirm tubing condition — for peristaltic, check the section under the rollers for ovalisation or surface damage.
4. Set the spray-rate profile: ramp-up (typically 20–30 % of nominal rate for 5–10 minutes, allowing the bed to reach steady state without initial over-wetting at the cold start), hold (validated nominal rate for the main spray phase), ramp-down (reduce to 30 % of nominal in the last 5–10 % of solution to avoid stranded over-wet zone at end-of-spray).
5. Trend actual flow vs setpoint continuously; investigate any deviation > 5 % sustained for > 30 s.
6. Cross-check flow meter against load-cell mass loss on the solution tank; mismatch beyond ± 3 % opens a deviation.
7. Monitor bed temperature and bed weight (where instrumented) for early signal of wet-equilibrium drift.
8. On completion, record total solution delivered, total batch mass change, and the integrated flow-vs-time trace in the batch record.

1. Single flow meter for a multi-nozzle line — per-nozzle imbalance is hidden, total flow looks fine, but one quadrant of the bed over-wets while another under-sprays.
2. Peristaltic tubing reused beyond validated cycles — rate drifts low (compression set in the tubing reduces stroke volume) and the operator chases endpoint with longer cycles rather than diagnosing the pump.
3. Constant spray rate from start to end — no ramp, frequent early over-wetting at the cold bed, sometimes recovered, sometimes not.
4. Blocked nozzles not detected — local under-spray plus neighbouring over-spray; the bed averages out to a passing PSD but with hot and cold spots that show up as content uniformity failures downstream.
5. No load-cell cross-check — flow meter drift or failure goes silent until residual solvent or LOD at end-of-dry reveals the discrepancy.
6. Atomisation pressure not linked to spray rate in the recipe — operator increases spray rate but atomisation stays at the same kPa, droplets get larger, the regime shifts from coating to granulation without anyone noticing.
7. Cleaning verification missed at the nozzle — residual binder from a previous batch carries forward, especially in shared facilities.
8. Spray-rate setpoint adjusted by operator without deviation — 'we always do this when the bed temperature drops' but the adjustment is not in the recipe and is not tracked.
9. End-of-spray declared on volume delivered rather than mass — for solutions with concentration drift across the tank (settling, supplier-to-supplier variability), volume-based delivery can be 5–10 % off mass.
10. Solution tank not stirred during long spray phases — concentration stratifies, the first half-tank delivers more solids than the second half-tank.

Per-nozzle spray rate scales approximately linearly with bed mass — a small-scale validation at 80 g/min per nozzle for a 30 kg bed gives a defensible starting point of 80 g/min per nozzle at production scale, provided the nozzle count scales with bed area. Total spray rate, by contrast, scales with bed area (≈ scale^2/3) and with evaporative capacity (which scales with inlet air flow, typically tracked to bed area). Scaling total spray rate linearly with bed mass — a naive 'multiply everything by the same factor' — leads to over-spray at production scale because evaporative capacity has not kept pace with mass.

The pragmatic scale-up rules: hold per-nozzle spray rate constant, scale nozzle count with bed area, hold atomisation pressure constant, scale inlet air flow with bed area, hold inlet air temperature constant. Confirm in PPQ at production scale; the receiving site PPQ batches almost always need a 5–15 % adjustment to per-nozzle rate to fit the local equipment characteristics (slight differences in plenum geometry, exhaust draw, manifold balance).

• Solid-dose pharma granulation and coating — universal CPP; the most-studied spray-rate workflows in industry.
• Effervescent coating — very tight spray-rate windows because the substrate is moisture-sensitive and any over-spray triggers reaction.
• Sustained-release pellets (Wurster) — slow controlled spray rate, often with multiple weight gains and intermediate cure steps.
• Veterinary chewable coating — palatant overcoats with controlled rates to lock flavour without altering core dissolution.
• Agrochemical water-dispersible granules (WG/WDG) — surfactant solution metering analogous to pharma granulation.
• Food encapsulation — flavour oils, essential oils, or actives sprayed at controlled rates onto carrier substrates.
• Industrial catalyst manufacture — precious-metal solution metered onto support pellets at strict rate to control loading.

• Pump calibration record bound as a phase precondition — calibration window expired = phase blocked.
• Per-nozzle flow capture, not aggregated total — imbalance surfaced live.
• Spray-rate profile locked in the recipe; flat profiles flagged at recipe approval.
• Load-cell mass loss reconciled against integrated pump flow; > ± 3 % opens a deviation.
• Atomisation pressure, inlet air temp and inlet humidity tracked alongside spray rate; recipe links the four CPPs.
• End-of-spray declared on mass delivered (with solution concentration evidence) — not volume.
• Trend dashboards per product per vessel: mean spray rate, cycle time, end-of-spray PSD; CPV alarms on drift.