Bench unit, pilot plant or something close to production, and the steam, chilled water, potable water, drainage and cleaning routine that decide whether a food technology teaching plant is still running in its second year.
A food and dairy pilot plant is the only teaching purchase in an engineering faculty where the equipment is the smaller half of the project. The stainless steel arrives in crates and is installed in a week. The steam, the chilled water, the potable water quality, the floor drainage, the extraction and the cleaning routine are a building project, and departments that treat them as a fit-out detail end up with a handsome plant that runs twice and then stops.
The first decision is scale, and it decides everything downstream. There are three honest positions.
Bench and single process units. A cream separator (DSNC), a laboratory pasteuriser (PASC), a yogurt incubator (IYDC), a curdling tank (CCDC), a cheese press (PVQC), a butter maker (EMANC) or a multipurpose processing vessel (VPMC). Each teaches one operation, sits on or near a bench, uses domestic quantities of product and can be cleaned by hand. This is where a department with no existing food services should start, and it is enough for a food science curriculum.
Pilot plants. A linked line such as the LE00/PM for pasteurised milk, LE00/Y for yogurt, LE00/CC for cheese and cottage cheese, LE00/CBI for cream, butter and ice cream, or LE00/MP for milk powder. On the food side the equivalents are pilot plants for juice (FR00/JP), jam (FR00/FJ), concentrated juice (FR00/JC), bread (CE00/B), pasta (CE00/P), milling and flour (CE00/MF), and product lines for tomatoes (TO00), citrus (CI00), vegetables (VE00) or cured and cooked meat products (CA00/CUPS, CA00/COPS). These behave like small factories. They need utilities, a technician and a cleaning regime.
Something close to production. A UHT unit (AUHT/CTS), a spray drying line for milk powder, a liquid packaging unit (EDLC) or a solids packaging unit (EDSC). At this scale the department is operating a small food plant, and the honest question is whether it wants to.
The mistake is buying at pilot scale with bench scale services. A pasteuriser that needs hot water at a temperature and flow rate will not hold its holding time on an undersized supply, and a student will record a passing result from a process that did not actually pasteurise anything. That is worse than not having the equipment.
A dairy technology syllabus in Bangladesh usually runs: reception and testing, separation, standardisation, homogenisation, heat treatment, culture and fermentation, and then a product route into yogurt, cheese, butter, ice cream or powder. Map the equipment to that sequence rather than buying by name.
Milk powder deserves its own paragraph because it is the item most often requested and least often understood. Spray drying needs a large volume of heated air, an exhaust route, cyclone or bag collection, and a plant room. It is a building services decision before it is a purchase order line. It also produces a hygroscopic powder that will cake in Dhaka humidity unless the packing area is dry, which means the product store is part of the specification, not an afterthought.
Departments teaching food technology rather than dairy specifically tend to want a fruit and vegetable processing line, a cereal line, or an oil line. Each is a different set of utilities. Fruit processing (juice, purees, jams, concentrates) is water, steam and acid; the plant is mostly stainless steel and the waste is high in sugar and organic load. Cereal processing (milling, bread, pasta, malting) is dust, flour and mechanical hazards, and it needs dust extraction rather than fume extraction. Oil production (AS00 for seed oil, AC00 for oil production, AEHC for hydrogenation, AEDC for deodorising) brings hot oil, vacuum and a fire risk profile that a general purpose laboratory is not built for.
Pick one line and equip it properly. A department with a partial fruit line, a partial cereal line and a partial dairy line has three plants that each lack one operation, and students who never see a product through from raw material to package.
These four are where the money goes that nobody put on the equipment list.
Steam or hot water. Many teaching units are self-contained with electric heating, which is the simpler route and puts a real electrical load onto a distribution board that may already be at its limit. Check the connected load of the whole plant against the existing supply before ordering, not after. Where a boiler is needed, an electric steam generator avoids a fuel supply and a chimney but is expensive to run continuously, so match it to the operating hours the plant will really see.
Chilled water. Pasteurisers and cooling sections need it, and a chiller is often the single largest utility item. Size it against simultaneous operation of the plant, and give it somewhere to reject heat that is not the process room.
Compressed air, oil free. Pneumatic valves, packaging machines and some transfer systems need it, and in a food plant it must be oil free and dried. This is not the same compressor as the one feeding the workshop.
Potable water, and its quality. The plant washes with the water you give it. Water carrying iron, which is common on Bangladeshi deep tube wells, stains stainless steel and discolours product. Hard water scales heat exchange surfaces and shortens the interval between descaling. Test the campus supply before commissioning and budget for softening or iron removal if the result requires it. This is a cheap test that prevents an expensive argument about who is responsible for a stained plant.
A food processing room is not a laboratory with equipment in it. The finishes are part of the process.
Add ventilation. Cooking, drying and pasteurising put moisture and odour into the air, and in a closed room in a Bangladeshi summer the condensation lands on the ceiling and drips onto the plant. Extraction over the wet operations, with make-up air, is a design item and not something to be added later with a wall fan.
Every pilot plant has to be cleaned after every run, and cleaning takes longer than the run. Whether the plant has an automated cleaning circuit or is cleaned manually, the department needs caustic and acid detergents, a place to store them, dosing equipment, protective equipment for the person doing it, and a written procedure with contact times.
More importantly it needs a person. A pilot plant with no dedicated technician silts up within one semester: a fouled heat exchanger, a seized valve, a culture contaminated by a badly cleaned vessel. The department then reports an equipment problem. Budget one technician's time against the plant before the tender goes out, and make it a named responsibility rather than an addition to someone's existing duties.
Train two people, not one. The technician you train will be promoted, transferred or will leave, and a plant whose only competent operator has gone is a plant that stops.
A teaching plant makes food, and somebody has to decide what happens to it before the first run rather than after. A university pilot plant is not a licensed food producer, and treating student output as saleable product creates a regulatory problem the department does not want. The workable positions are internal consumption under a written departmental rule with a documented process record, or disposal as waste.
If it is disposal, understand what you are disposing of. Dairy and food waste has a very high organic load. Whey, wash water and spilled product will overload a septic tank and will produce odour and a complaint from the building next door. Where the plant runs regularly, the honest answer is a small balancing and treatment provision, which has the useful side effect of being a teaching asset for the environmental engineering course.
Do not accept a power-on test as commissioning. Specify a supervised first production run: real raw material, the full process, the product made, the plant cleaned afterwards, with the department technician operating and faculty present. Hold a meaningful part of the payment against that run rather than against delivery.
Ask for three documents at the same time: the cleaning procedure with chemical concentrations and contact times, the annual consumables and spares list with part numbers, and the process record sheet the plant is expected to produce. Those three pieces of paper are what let a new technician run the plant in year three, and they are much easier to obtain while the commissioning engineer is still on site.