Agitated Nutsche Filter Dryer (ANFD): A Buyer's Guide

An agitated nutsche filter dryer (ANFD) is a closed, jacketed vessel with a flat filter plate in its base and an agitator that turns and travels up and down through the cake. In that one vessel it filters a slurry, washes the cake, reslurries it if needed, dries it and discharges the powder, so the product never moves between machines [1, 2]. A 2022 study by the University of Leeds and AbbVie calls agitated filter bed drying the pharmaceutical industry's preferred way to isolate active pharmaceutical ingredients (APIs) [3].

To buy one well, fix the cake volume per batch first; filter area and cake depth follow from it.

Why this matters now

Most ANFD options are fixed at order: heated agitator, discharge isolator, cleaning nozzles, filter plate and type of medium. Containment claims differ too: De Dietrich's pack off station comes in OEB4 or OEB5 versions [2], 3V Tech offers isolators up to OEB6 [4], and 3M's suggested banding table stops at OEB 5 [5]. Two bids that both say "OEB 5" may not promise the same thing, so the RFQ (request for quotation) must define the terms.

What an ANFD does in one batch

A nutsche is a flat bed filter: gas pressure or vacuum drives the liquid through a horizontal filter medium. The ANFD adds a jacket, an agitator and a sealed discharge, and works under both vacuum and pressure [1].

Filtration. As the mother liquor, the liquid the crystals formed in, drains, a cake builds up. PSL's agitator can run in either direction, levelling the cake one way and turning it over gently the other [1]. Levelling closes cracks that let gas or solvent bypass the cake.

Washing. Solvent is pushed straight through the cake, or the agitator resuspends the cake in solvent, a reslurry wash, before the liquid is pressed out [1].

Drying. Heat enters through the jacketed shell and, on many units, the filter plate and agitator; De Dietrich claims 15% to 20% faster drying with a heated shaft and agitator [2]. A filtered cake typically holds about 10% to 40% solvent, and drying can take hours to days [3].

Discharge. The agitator ploughs the powder to a side valve that seals against pressure and vacuum [1, 2]. What remains is the heel, a thin layer below the blade tips. De Dietrich pushes it out with nitrogen, quoting up to 99% efficiency [2].

Agitation and particle breakage

The Leeds study says blade shear can break crystals or cause agglomeration; typical industrial speeds are about 5 to 10 rpm, and as low as 2 rpm for fragile material [3]. In their laboratory tests a smaller gap between blades and base generally caused more breakage, and they warn against agitating a cake that is still wet [3]. Ask bidders for the speed range, lift control and minimum clearance, and trial your product first.

Sizing: batch volume, cake depth and filter area

Sizing starts with the solids, not the slurry. Lab filtration tests give the cake loading, kilograms of dry product per litre of wet cake. Then:

• cake volume per batch = dry product per batch ÷ cake loading
• cake depth = cake volume ÷ filter area

PSL's published sizes show the pattern [1]:

A deeper cake slows filtration, washing and drying; a very shallow one gives the blades little to work on. The Leeds authors say a bed of two to three times the blade height is normal [3]. Check slurry and wash volumes against liquid capacity too.

Materials, pressure rating and containment

Four makers' published data differ widely:

Materials, media and ratings

Makers offer 316L stainless steel or nickel alloys such as alloy C22 for wetted parts [1, 4]. Where even Hastelloy C22, C276 or B3 is not resistant enough, De Dietrich and GMM Pfaudler offer glass lined units [2, 6]; De Dietrich presents its glass lined filter dryer as a lower investment alternative to an alloy C22 unit [7]. Glass has its own limits; see our guide to jacketed vessels and glass lined reactors.

PSL offers polypropylene, PEEK or PTFE cloth as a consumable, or sintered metal mesh, which costs more but is cleaned in situ [1]. State the internal finish as an Ra value; none of these maker pages gives one.

Rate each heating circuit as well as the vessel, and state the code and marking you need; in 3V Tech's range, minus 1 barg is full vacuum [4].

Containment, OEB and cleaning

An occupational exposure band (OEB) groups compounds by potency; each covers a range of occupational exposure limits (OELs), the airborne concentrations workers may be exposed to. In 3M's suggested table, OEB 3 covers OELs above 10 and up to 100 µg/m³, OEB 4 above 1 and up to 10 µg/m³, and OEB 5 below 1 µg/m³; 3M warns its tables may not suit every company [5]. So put the OEL in µg/m³ in the URS, with the containment test method. ISPE's SMEPAC guide (third edition, December 2024) sets out standard tests that run a surrogate powder through the equipment and measure what reaches the air and what settles nearby [8].

With the vessel closed, the weak points are the seals and the discharge. Cleaning must reach the dust filter too; De Dietrich's glass lined unit has a WIP (wash in place) device for both [7].

Flammable solvent is not the only hazard: the dried powder may form an explosive dust cloud. HSE lists dusts as well as solvents among the substances covered by DSEAR, the Dangerous Substances and Explosive Atmospheres Regulations 2002, under which employers must manage fire and explosion risk and classify places where explosive atmospheres can arise [9]. State the area classification for drives, instruments, discharge and dust filter, any nitrogen inerting, and give bidders the product's dust explosibility data. This is general information, not legal advice.

What the URS, RFQ and FAT must cover

Write the URS (user requirement specification) in process terms and let bidders propose the machine; our URS template gives the structure.

Our FAT checklist covers the factory acceptance test method.

Practical example

Take an illustrative intermediate: 150 kg of dry product per batch, lab cake loading 0.5 kg per litre. The cake is 150 ÷ 0.5 = 300 L, or 360 L with 20% allowed for growth, and the slurry is 750 L. Both bidders quote the 400 mm maximum cake height and liquid capacities in PSL's table [1].

Bid A holds at most 0.6 m² × 0.4 m = 240 L of cake, or 120 kg of product, so every batch needs two loads. A lower price is no saving.

For Bid B, a reslurry wash at an illustrative 2 L of solvent per kg is 300 L; with 300 L of cake that makes 600 L, within 1,000 L. A 300 mm cake suits blades about 100 mm to 150 mm high [3]. At an illustrative 25% solvent by mass, the wet cake weighs 150 ÷ 0.75 = 200 kg and holds 50 kg of solvent; ask each bidder for the expected drying time and its basis.

Find makers who can build it

Supplier Discovery searches the web for filter dryer makers from a plain text description, pasted RFQ or uploaded spec. It extracts makers' published parameters with source links and gives each a fit verdict, from Strong to Rejected. Certifications are shown as claimed, not verified, nothing is sent automatically, and you can try three searches free without signing up.

Ready to find ANFD makers that match your URS? Start a free 14 day ChemCapital trial, with no credit card needed.