Ají amarillo paste, ají panca, manjar blanco and lúcuma jam are all batch products, and all four are cooked in the same kind of vessel. What separates a good batch from a rejected one is not the recipe. It is the jacket temperature, the scraper clearance and the altitude of the plant.
Peru cooks in batches, and the batch is where quality is won or lost
Walk into a Peruvian food plant and the cooking room usually has three or four kettles of different sizes, one of them older than the supervisor. The complaints are also consistent: the manjar blanco catches at the bottom in the last twenty minutes, the ají paste darkens at the rim, the lúcuma jam loses its colour, and nobody can say what the jacket pressure actually was. This article is about the engineering behind those four complaints, written from the batch sheets and failure reports we collect from customers in Lima, Arequipa, Trujillo and Cusco.
1. The four products that pay for the kettle
| Product | Typical batch | Process | Target | Main risk |
|---|
| Ají amarillo paste | 300–500 kg | 90–95 °C, 40–50 min, stirring 40–60 rpm | pH ≤ 4.2, 20–24 Brix | Carotenoid oxidation and scorching at the wall |
| Ají panca paste | 300–400 kg | 90–95 °C, 50–60 min | 18–22 Brix | Seed and skin separation, uneven colour |
| Manjar blanco | 200–400 kg | 100–104 °C, 90–120 min, 25–35 rpm | 68–72 Brix | Burn-on, over-dark Maillard reaction |
| Mermelada de lúcuma or aguaymanto | 250–400 kg | 85–90 °C, pectin added at 50–60 °C | 65–67 Brix | Pectin degradation if held above 95 °C |
| Fruit pulp concentrate | 500 kg | 70–80 °C, preferably under vacuum | 28–32 Brix | Aroma loss and browning |
The table explains why one kettle cannot be optimised for all five. Manjar blanco needs high temperature and continuous scraping at low speed; ají paste needs moderate temperature and faster agitation to avoid a stagnant layer; fruit pulp wants the lowest temperature that still gives a shelf-stable product. A plant that runs all three in one vessel is compromising at least two of them, and usually knows it.
2. Steam jacket, electric or direct fire: what actually changes
Steam jacket. Saturated steam at 0.3 MPa is about 143 °C, and the jacket gives a large, even heat-transfer area with no local hot spot. This is the only configuration we recommend for manjar blanco above 200 kg per batch. The jacket is designed at 0.45 MPa, hydro-tested at 1.5 times design pressure for 30 minutes, and fitted with a safety valve set at 0.35 MPa. It needs a boiler: a 600 L kettle at full load pulls roughly 250–300 kg/h of steam, which is why plants without a boiler should not be sold a steam unit.
Electric and electromagnetic. Our 200-A frame is a 20 kW electromagnetic kettle with a 1.5 kW variable-frequency stirrer, 800 mm inner pot diameter, 550 mm depth, 8 mm pot thickness (6 + 2 mm) and a rated capacity around 1,000 kg/h. It suits plants with no boiler and batches up to a few hundred kilos. At 380 V three-phase, 20 kW draws about 30 A per phase, so the existing panel and cable size must be checked before delivery; we have seen a 20 kW kettle installed on a 4 mm² feed meant for a 7.5 kW mixer.
Direct gas. Lowest capital cost and the reason so many small Peruvian workshops still use it. The flame side reaches 800–1,000 °C, and even with a diffuser plate the wall temperature overshoots during the last stage of a manjar blanco boil. If a plant insists on gas, insist on a double-bottom pot and a scraper, and accept a darker finished colour.
Selection rule we use. Under 200 L and no boiler: electric. Over 300 L with a boiler: steam jacket. Manjar blanco at any size: steam or electric with a spring-loaded scraper, jacket temperature capped near 130 °C, never a bare flame.
3. Geometry and agitation: the details that stop burn-on
Burn-on is a heat-transfer problem before it is a cleaning problem. Three numbers control it:
- Scraper clearance. 3–5 mm between the PTFE blade and the pot wall, spring-loaded so the blade follows a slightly out-of-round pot. Beyond 5 mm a stagnant film forms, and that film reaches 120 °C+ while the product bulk is still at 95 °C. Blade wear is measurable: we ask plants to check clearance every 500 hours and replace at 6 mm.
- Agitation pattern. A planetary head with a counter-rotating anchor covers the full hemisphere; a single central paddle leaves an annulus at the bottom that no scraper touches. For manjar blanco, 25–35 rpm; for ají paste, 40–60 rpm; for jam, 30–45 rpm after the pectin is in.
- Jacket coverage and headspace. Jacket on at least 60 percent of the shell, 15–20 percent headspace for foaming products, and 50 mm rock wool insulation under stainless cladding so the operator is not standing next to a 140 °C wall.
Temperature sensing deserves its own sentence. A probe welded into the jacket reads steam temperature, not product temperature. The product probe belongs 50 mm above the bottom, and the uniformity should be verified by mapping three points (centre, wall, mid-depth) during commissioning; ±2 °C across the batch is achievable on a well-baffled kettle and is not achievable on a kettle with a single probe and no agitation.
4. Materials, hygiene and the Peruvian inspection
Product contact is AISI 304, 2B finish, Ra ≤ 0.8 µm. Where the product carries more than about 3 percent salt — rocoto sauces, brine-based marinades — we quote 316L, because 304 pits at the weld-affected zone in chloride service and the pit is invisible until the inspector finds it. Welds are ground, pickled and passivated; no threaded fittings in contact with product; the outlet valve is a clamped sanitary type, not a ball valve with a cavity.
The regulatory frame is the same one that governs the rest of the plant: DS 007-98-SA and the plant's HACCP plan, with DIGESA oversight. The equipment's contribution to that file is documentation: material certificates, food-contact declarations for gaskets (EPDM for steam and water, silicone where fat resistance is not an issue), a cleaning procedure with validated parameters, and a maintenance log. We supply all of it in Spanish, because the inspector will read it in Spanish.
Cleaning in practice: 1.5–2 m/s caustic at 75–80 °C for 20–30 minutes, rinse to neutral, acid cycle weekly where hard water is used, and ATP verification at four points. A kettle that is cleaned by hosing down and wiping is a kettle that will fail a swab test within a month.
5. Utilities and site conditions: the checklist before the kettle lands
- Steam. 250–300 kg/h at 0.4 MPa for a 600 L kettle, DN25–32 supply, thermodynamic trap, filter, pressure-reducing valve and a safety valve on the jacket. A jacket with a failed trap fills with condensate and heats at half rate; this is the most common "the kettle is slow" call we get.
- Power. 220/380 V 60 Hz. The 20 kW electric frame needs a 40 A breaker, 4 * 6 mm² copper and a 30 mA RCD, with earthing measured at commissioning. At 60 Hz the stirrer runs 20 percent faster than the 50 Hz nameplate; the VFD must be set to a 60 Hz base so the scraper speed in the recipe book still means what it says.
- Water hardness. Lima water runs roughly 150–300 mg/L as CaCO₃. A 0.5 mm scale layer costs 5–8 percent of heat transfer and is the second most common cause of slow batches. A softener upstream of the jacket make-up is cheaper than descaling twice a year.
- Floor and access. Plan for the footprint (the 200-A is 3000 * 1500 * 2000 mm), a 500 kg/m² floor load, slope to drain, and 100 mm clearance around the frame for cleaning. Tilting kettles need another 1,200 mm in front for the trolley.
- Altitude. This one is specific to Peru and frequently missed. Water boils at about 89 °C in Cusco and 90 °C in Arequipa's higher districts. An open kettle there cannot reach the 100–104 °C that manjar blanco needs, no matter what the thermostat says. At altitude, either specify a pressure kettle or move to a vacuum evaporator and adjust the recipe for the lower boiling point.
- Coastal humidity. In Lima and further north, IP55 motors, an anti-condensation heater and passivated stainless are not optional. We have replaced control cabinets in Iquitos that failed in eighteen months because condensate ran down the inside of an unheated enclosure.
6. Fault table from our service reports
| Symptom | Probable cause | Field fix |
|---|
| Burn-on at the bottom within 20 minutes | Scraper clearance above 5 mm or worn PTFE blade | Reset to 3–4 mm, replace blade, check spring tension |
| Slow heat-up, jacket hot but product cold | Air lock or failed steam trap, or 1 mm scale | Vent the jacket, replace the trap, descale |
| Uneven colour from centre to wall | Single-point control, agitator too slow | Three-point mapping, raise to 35 rpm, add baffle |
| Steam leak at the jacket gasket | Over-pressure or thermal cycling | Verify the 0.35 MPa safety valve, retorque to spec, new gasket |
| Product left in the pot after tilting | Tilt angle below 100°, or bottom valve with a cavity | Specify 105–110° tilt, or a clamped sanitary valve |
7. Buying and shipping to Peru
Orders ship FOB Qingdao or CFR Callao, occasionally Matarani for southern plants. Payment is commonly 30 percent T/T with the order and 70 percent before shipment, or an L/C at sight. Manufacturing takes 25–35 days, the sea leg to Callao 30–38 days, and clearance through SUNAT about a week with the importer's RUC and a broker.
We recommend a factory acceptance test with a simulated product — water thickened with 2 percent CMC to approximate the viscosity of a heavy ají paste — and a video record if the buyer is not travelling. It costs one day and catches the scraper and jacket problems before the container is sealed. Spare parts to keep on the shelf: one set of scraper blades, two jacket gaskets, one steam trap, one product-probe PT100. That list is worth less than a single emergency air freight.
Questions from Peruvian food plants
Can one kettle do ají paste and manjar blanco? Physically yes, but you will compromise one of them. If the batch sizes are similar, specify two speeds and a jacket temperature controller that can hold 130 °C; if manjar blanco is the main product, do not buy a gas-fired pot.
Do we need 316L? Only for high-salt products such as rocoto or brine marinades. Standard ají pastes and jams are fine in 304, provided the welds are passivated.
How long does a batch take on the 200-A? About 45–60 minutes to bring 300 kg of ají paste from 20 °C to 95 °C, plus holding time. Manjar blanco is a 90–120 minute process because of evaporation, not because of heating power.
What about CIP? The kettle is cleaned in place for the pot and manually for the head and scraper arm. We supply the spray ball option if the plant wants a fully closed cleaning cycle.
Your message must be between 20-3,000 characters!
