Skip to content
PPXLROASTERS
PXL Roasters
  • History
    • Process

    Thermal shock

    Directed fermentations.

    THE PROCESS IN MOTION

    What does thermal shock really show?

    Start the scene or choose a stage. The animation simplifies the mechanism; what follows gives the details and the available references.

    Fermenting as whole cherry

    The lot documented by the importer starts from 95% ripe cherries, washed with water filtered by reverse osmosis and then by ultraviolet light, so that the added starter dominates. First fermentation of 48 hours at 18 °C, in a closed bioreactor with a valve: the valve lets the carbon dioxide out, it does not create a vacuum.

    Read the stages without animation
    1. Fermenting as whole cherryThe lot documented by the importer starts from 95% ripe cherries, washed with water filtered by reverse osmosis and then by ultraviolet light, so that the added starter dominates. First fermentation of 48 hours at 18 °C, in a closed bioreactor with a valve: the valve lets the carbon dioxide out, it does not create a vacuum.
    2. Pulping, re-fermenting in rotationThe coffee is pulped; the honeys, the sweet juice of the fruit, are kept and re-inoculated. Second fermentation of 48 hours at 21 °C, starter added to the mucilage, in a rotating fermenter: the rotation homogenizes the mass and limits the gap between top and bottom.
    3. The shock: bringing the liquid to 40 °CAt the end of the tank stage, the fermentation liquid is heated to around 40 °C with the beans in it; a roaster gives 40 to 50 °C. At 40 °C, no microorganism in the tank dies: their growth ceiling is exceeded, yeasts die only above 50 °C. If the step goes on, degradation continues. Its duration is published nowhere.
    4. Washing at once in cold waterThe coffee is washed without delay in cold water, at 12 °C on this lot. Cold slows the microorganisms without stopping them: refrigeration keeps café en baba, the pulped bean still covered with mucilage, only below 8 °C, and lactic acid bacteria still reproduce at 0 °C. The idea of pores dilated and then closed remains a practitioner’s hypothesis.
    5. Drying in a machineMechanical drying for 34 hours at 35 °C, in air at 25% relative humidity, down to 10 to 11% bean moisture. A cold, water-logged bean left waiting becomes covered with condensation and molds.

    To remember: Thermal shock ferments nothing: it ends a fermentation, and neither the hot bath nor the cold bath stops the microorganisms dead.

    In history: The process is associated with Finca El Paraíso of Diego Samuel Bermúdez Tapia, in Cauca; the first lot sheets that mention it date from the 2021 harvest.

    Read the full article and its sources ↓
    Several closed stainless steel fermentation tanks, fitted with red-handwheel valves and pressure gauges, lined up in the room of a wine cellar; the image shows winemaking equipment, but closed tanks with valves and gauges are of the same type as those used for coffee fermentation
    Stainless steel fermentation tanks lined up with valves and gaugesDrew Cuddy · CC BY-SA 2.0
    FamilyDirected fermentations
    Also calledthermoshock, thermal shock washing, choque térmico, bean sealing, thermal shock fermentation (a false name: the step ferments nothing, it ends a fermentation run before it)

    Where it is practiced

    The process is associated with a farm, Finca El Paraíso, in Piendamó, in the department of Cauca, Colombia. It is run by Diego Samuel Bermúdez Tapia. The Taza de Excelencia 2018 competition lists the farm under this name, in Piendamó, Cauca. The importer who sells these lots gives 1,930 meters for the farm. Its sheet for the lot documented below gives 1,750 meters for the plot. No source read publishes an altitude range for the farm. See GRAY AREA.

    The practice spread beyond this farm. Josué David Valencia describes a comparable step at the finca Pachamama, his family farm, whose department the article citing him does not name. Another farm in Cauca, which its owner Wilton Benítez himself calls “Finca El Paraíso-92”, publishes a bean sealing by impacts of hot water and cold water. The two farms are distinct. The date of development is established by no source. The earliest importer lot sheets bearing the mention “Thermal Shock” that can be consulted concern the 2021 harvest, analyzed in January 2022.

    The principle

    Thermal shock is not a fermentation. It is a way of ending one. At the end of the tank stage, the fermentation liquid is heated to around 40 °C, with the beans in it. The coffee is then washed without delay in cold water, at around 10 to 12 °C.

    The explanation given by the producer is mechanical. Heat would dilate the pores and membranes of the bean. The bean would then absorb the aroma precursors from the medium. The sudden cold would close everything again and fix these compounds in the kernel. This explanation is a practitioner’s hypothesis, never measured.

    What the literature allows us to say about the hot bath is more cautious. At 40 °C, no microorganism in the tank dies. Cenicafé gives three growth ranges for the flora of the mucilage, the layer of sweet flesh stuck to the parchment.

    • Yeasts grow from 5 to 39 °C, with an optimum of 28 to 35 °C, and all die above 50 °C.
    • Lactic acid bacteria grow from 25 to 30 °C, but still reproduce at 0 °C.
    • Enteric bacteria develop from 22 to 37 °C.

    At 40 °C, then, one goes above the growth ceiling of the yeasts as well as that of the enteric bacteria. One remains very far from the temperature that kills them. The same source shows that fermentation goes faster when the ambient temperature rises. It measured this between refrigeration and the Colombian ambient, not at 40 °C. Extending the rule to the hot bath remains an extrapolation. What is established is that 40 °C sterilizes nothing.

    The cold bath, for its part, acts in the opposite direction, and it is the defensible part of the process. Cold slows the microorganisms and the enzymes. Cenicafé put a figure on this for the lipids of the mucilage. At 20 hours of fermentation, nearly 40% of these lipids break down at 20.5 °C, against 16% under refrigeration at 6.6 °C. Refrigeration truly preserves café en baba, that is, the pulped bean still covered with its mucilage, only below 8 °C. The 10 to 12 °C of the rinse remain above this threshold.

    The sequence

    The protocol below is the one published by the importer for a Castillo “Red Plum” lot from Finca El Paraíso. It varies from one lot to another on the farm.

    • Sorted harvest, 95% ripe cherries and 5% half-ripe. Ripeness governs the sugar available, and therefore what the fermentation will have to consume.
    • First wash with filtered water, reverse osmosis and ultraviolet C. The aim is to lower the microbial load brought by the fruit, so that the added starter dominates.
    • First fermentation, known as anaerobic, 48 hours at 18 °C, in a closed bioreactor with a pressure relief valve, with selected microorganisms. At 18 °C, one is below the yeast optimum: fermentation is slow and the operator stays in control.
    • Pulping. The honeys, the sweet juice from the fruit, are kept and re-inoculated.
    • Second fermentation, 48 hours at 21 °C, starter added to the mucilage in a rotating fermenter. The rotation homogenizes the mass, which limits the gap between the top and bottom of the tank.
    • Thermal shock. The fermentation liquid, beans included, is brought to 40 °C. The coffee is then washed in cold water, at 12 °C.
    • Mechanical drying, 34 hours at 35 °C, air held at 25% relative humidity, down to 10 to 11% bean moisture.

    The word “anaerobic” is misleading here, as almost everywhere in coffee. Nobody drives the air out of the tank. It is closed, the microorganisms consume the oxygen present, and the medium becomes anoxic by depletion. The valve serves to let out the carbon dioxide produced, not to create a vacuum.

    On drying, a word of caution. The only controlled study read on air temperature compares 40 °C, 50 °C and solar drying. Germination drops there at 50 °C, not at 40 °C. The 35 °C of the protocol is below both setpoints tested. No precise threshold is established for all that.

    A roaster who buys these lots publishes a wider range. This roaster gives 104 to 122 degrees Fahrenheit for the hot bath, that is, 40 to 50 °C. It gives about 50 degrees Fahrenheit for the rinse, that is, about 10 °C.

    To place the gap, here is the Colombian reference practice on an ordinary washed process. Fermentation is stopped when the pH of the mucilage is between 3.7 and 4.1. The coffee is washed with clean water at ambient temperature, 1.7 to 2.0 liters per kilogram of café en baba, in four successive rinses. No hot step appears in it.

    What it gives in the cup

    What is measured. An importer’s analysis sheet, drawn up in Amsterdam in January 2022, gives three cuppings of the same lot. It is a Castillo “Red Plum”, thermal shock and washed process, grown at 1,750 meters. The scores are 87.25, 87.25 and 87.00, an average of 87.17 points, for a bean moisture of 10.4%. The descriptors noted are strawberry, candy, tropical fruit, carambola, peach, lemonade and milk chocolate. This measurement is real and traceable. It has, however, neither a control nor blinding, and the panel belongs to the seller.

    What is not measured. No publication isolates the effect of thermal shock. A search in Europe PMC and in Crossref in September 2026 returns no work on the subject.

    The closest point of comparison is a controlled Cenicafé study on fermentation temperature. It covers three varieties, Castillo, Cenicafé 1 and Tabi, at two setpoints, 15 and 30 °C. A spontaneous fermentation serves as control. Temperature changes the kinetics. At 15 °C, the breakdown of the mucilage lengthens by more than 24 hours. The final pH falls below 3.5 there, against about 4.0 at 30 °C. It does not change the score. The gap between the two setpoints is not significant, and neither is the gap with the spontaneous fermentation.

    The other useful point comes from the measurement of aroma transfer. Three labeled compounds were followed from a liquid medium to the bean, at 25 °C, at 0, 6, 12, 24 and 48 hours. In the medium containing pulped beans and yeast, at 12 hours, the values are as follows.

    • 2-phenylethanol: 11.2 µg/g of coffee.
    • Isoamyl acetate: 1.3 µg/g.
    • Butanal: 0.2 µg/g.

    The parchment clearly slows the passage of 2-phenylethanol. Transfer therefore exists, but it is counted in hours. This weighs on the mechanism claimed. The only duration indication published on thermal shock comes from a trade press article, which speaks of a few seconds. At that scale, the transfer measured above has no time to take place.

    What is claimed. The trade attributes to thermal shock a fruity, candied intensity. A trade press investigation published on July 8, 2026 qualifies this talk. The producer interviewed, Josué David Valencia, estimates that about 90% of the co-fermented lots he comes across mention thermal shock in their traceability. He attributes the real change in taste to four factors.

    • The added yeast.
    • The other ingredients rich in glucose.
    • The duration of the fermentation.
    • The medium in which it takes place.

    Temperature is not on his list. A roaster who sells these coffees even writes that scientific research would be needed to establish the origin of the effect.

    What goes wrong

    The main risk is in the hot bath, and it is counterintuitive. Bringing the tank to 40 °C does not stop it. No flora dies there, since yeasts die only above 50 °C. If the step goes on, degradation continues at the very moment one thinks one is finishing. The defect that follows is the vinegary bean (vinagre, in Spanish). Cenicafé links it to the prolonged stay of the beans in the fermentation acids, above all acetic acid. In the cup, this gives sour and fermented flavors (fermento, in Spanish).

    The 40 °C also exceeds the growth ceilings of the tank. Yeasts stop at 39 °C and enteric bacteria at 37 °C. The whole flora is thus pushed out of its growth range just before washing. It is not known which one starts up again first when the temperature falls back.

    The cold bath stops less well than advertised. At 10 or 12 °C, one stays above the 8 °C below which refrigeration really preserves café en baba. Lactic acid bacteria still reproduce at 0 °C. The stop is a slowdown, not a cut-off.

    Next come the ordinary risks, made worse by the step.

    • Delay in drying. The bean comes out cold and water-logged. If it waits, condensation and molds set in, and the moldy defect follows.
    • Dirty water. The hot bath takes place in the fermentation liquid itself, which concentrates everything the tank contained.
    • No thermometer. Without measurement, water said to be at 40 °C can be much hotter, and part of the flora is then killed and the surface of the bean is cooked.
    • Unverifiable mention on the bag. No duration, no temperature and no threshold is standardized. “Thermal shock” on a label guarantees no specific practice.

    Benchmarks

    • Advertised hot bath40 °C on the documented lot, 40 to 50 °C according to a roaster
    • Advertised cold bath12 °C on the documented lot, about 10 °C according to this roaster, about 14 °C on another farm
    • Duration of the two bathsnot published as an instruction; the trade press speaks of a few seconds
    • Yeast growth5 to 39 °C, optimum 28 to 35 °C, mortality above 50 °C
    • Enteric bacteria growth22 to 37 °C
    • Lactic acid bacteriagrowth from 25 to 30 °C, reproduction still possible at 0 °C
    • Cold that truly preserves café en bababelow 8 °C
    • End of fermentation in the Colombian washed processmucilage pH of 3.7 to 4.1
    • Reference wash1.7 to 2.0 liters per kilogram of café en baba, in four rinses
    • Drying of the typical lot34 hours at 35 °C, air at 25% relative humidity, down to 10 to 11%
    • Score measured on an importer’s lot, three cuppings, no control87.17 points
    Gray areaThe duration of the two baths is not published as an instruction. This is the most important point of the process, since it decides whether the heat acts for a second or an hour. Neither the producer, nor the importer, nor the roaster consulted gives it. Only one indication exists, in a trade press article of July 2026: the change would take place in a few seconds. This is a description of the principle, not an instruction measured on a tank. Temperatures diverge by source. - Hot bath: 40 °C at the importer, 40 to 50 °C at a roaster. - Cold bath: 12 °C at the importer, about 10 °C at this roaster, about 14 °C on another farm. - A value of 85 °C for thirty seconds circulates on the web, with no identifiable source. This last value could not be linked to any producer document. Wilton Benítez does publish a bean sealing by hot water then cold water on his own farm in Cauca. He gives no temperature for it. Any high value attributed to him is therefore unverifiable as it stands. The altitude of the farm is not established. The importer writes 1,930 meters for the farm. Its lot sheet writes 1,750 meters for the plot. No source read gives a range, and the two figures are not enough to build one. The attribution is not clean. The process goes back to Diego Samuel Bermúdez Tapia. The name “Alex Bermúdez” nevertheless circulates in several online glossaries. It is in all likelihood his brother Alexander, co-founder of the farm in 2008. The confusion could not be settled. A roaster also writes that Bermúdez and Benítez developed the process together, without any independent source confirming it. The confusion of names has a material cause. Wilton Benítez calls his own farm “Finca El Paraíso-92”. It is distinct from the Finca El Paraíso of Diego Samuel Bermúdez Tapia. Both are in Cauca and publish a hot water then cold water pass. Benítez’s site does not name his municipality, only the vereda Los Arados. The spread beyond this farm is poorly documented. Josué David Valencia describes a comparable step at the finca Pachamama, his family farm. The article citing him does not name the department. On this farm, the 40 °C is reached by microbial activity itself, not by external heating. The practice is therefore not the same as at El Paraíso. The date of invention is not established. The farm’s reputation dates from 2018. That year, a Bourbon lot scored 89.76 points and finished tenth in the Colombian Taza de Excelencia. It sold for 54.10 dollars a pound. The competition records this lot as “Double Anaerobic”, not as a thermal shock. The mechanism is not documented. No measurement of the porosity or permeability of green coffee has been published, before and after treatment. No trial compares a thermal shock lot with its own control on the same harvest. Finally, a methodological caution. Many web pages describing this process copy one another. One of them claims that a 2021 world barista champion competed with a lot from El Paraíso. The 2021 title does go to Diego Campos, for Colombia. No source consulted says which coffee he competed with, and none ties him to El Paraíso. None of these pages is cited here. == The dilation of pores now circulates in a Springer book This sheet writes that the explanation of thermal shock by the dilation and then closing of the pores of the tissue is a practitioner’s hypothesis, never measured. A scientific book published in 2025 takes it up as it stands, and this fact deserves to be added to the file, not because it validates the hypothesis, but because it shows how an explanation spreads. Hackler and Särchen describe the process in purely affirmative terms: a temperature shock is used during fermentation to dilate and then rapidly close the pores of the tissue layers of the cherry, and a higher temperature serves to make aroma compounds penetrate the bean, the tank being for example heated to forty degrees before a rapid drop. They mention the use of hot water and cold water in turn, and specify that the process is followed by the usual processing and drying. The disagreement with this sheet does not concern the protocol, on which the two texts agree in detail, including on the value of forty degrees. It concerns the status of the explanation. This sheet files it among the unmeasured hypotheses; the book writes it in the indicative, as a known mechanism. Checking the source is decisive here, and easy to do: the only reference the authors give for this passage is a professional popularization page, Barista Hustle, accessed on November 14, 2024. It is not a peer-reviewed article, it is not a measurement, and it is precisely the type of source that this sheet says keeps the explanation circulating. That a claim is printed by a scientific publisher does not change the nature of what supports it. The interest of this case is therefore twofold, and it deserves to be written this way. On the one hand it justifies the caution of this sheet by making it concrete: here is a practitioner’s hypothesis that, in a few years, has passed from a professional site to a Springer book, and that will henceforth be cited with the authority of the latter. On the other hand it provides one more source to cite, not to support the explanation, but to document its spread. What would make it possible to settle the matter remains what this sheet already asks for, and it must be repeated because the measurement is not out of reach: a direct observation of the structure of the tissues before and after a hot then cold cycle, by microscopy, and a comparative assay of a marker compound in the bean, with a control run at constant temperature on the same lot. Two of the three elements of the supposed mechanism, the dilation of the pores and the penetration of aroma compounds, are observable. No source consulted, neither this book nor the professional site from which it draws its information, reports this observation. Sources for this passage: Hackler and Särchen, chapter 3, p. 34, whose only source for this passage is Barista Hustle, Deconstructed Fermentation, accessed on November 14, 2024.
    Finca El Paraíso, “Authenticity” page of the producer’s site, accessed in September 2026, sections on thermal shock, condensation drying, yeast biocatalysis and antimaceration (no figures given); The Coffee Quest, producer profile “Diego Bermudez from Cauca, Colombia”, detailed protocols of the Red Plum, Candy Geisha, Lime, Lychee, Rose Tea and Manjar Blanco lots, and altitude of 1,930 meters; The Coffee Quest, producer profile “Diego Samuel Bermudez, from Finca El Paraiso in Cauca”, for Piendamó and the year 2008; The Coffee Quest, lot sheet CL21091 / SG-1767, Finca El Paraíso, Castillo Red Plum, Thermal Shock Washed, sensory analysis Cropster QC-4817 of January 17, 2022 and physical analysis QC-4806 of January 14, 2022, Amsterdam; Black & White Coffee Roasters, Trevor Clark, “What Does Thermal Shock Process Mean?”, August 2024, for the range of 104 to 122 °F, the rinse at about 50 °F, the joint attribution to Bermúdez and Benítez and the admission that scientific research remains necessary; MTPak Coffee, Dale Erwin, “Thermal shock coffee process: how temperature manipulation affects green coffee quality”, July 8, 2026, interview with Josué David Valencia, finca Pachamama; Granja El Paraíso-92, Wilton Benítez’s site, home page and “About us”, accessed in September 2026, for the bean sealing with hot water and cold water, with no temperature published; Puerta Quintero G.I., “Factores, procesos y controles en la fermentación del café”, Avance Técnico Cenicafé no. 422, August 2012, for the temperature ranges of the microorganisms, the breakdown of the mucilage lipids, the refrigeration threshold, the stopping pH, the wash volume and the origin of the vinegary bean; Peñuela-Martínez A.E., García-Duque J.F., Sanz-Uribe J.R., “Characterization of Fermentations with Controlled Temperature with Three Varieties of Coffee (Coffea arabica L.)”, Fermentation 9(11):976, 2023, DOI 10.3390/fermentation9110976; Peñuela-Martínez A.E., Sanz-Uribe J.R., Medina-Rivera R.D., “Influence of drying air temperature on coffee quality during storage”, Revista Facultad Nacional de Agronomía Medellín 76(3):10493-10503, 2023, DOI 10.15446/rfnam.v76n3.104115; Hadj Salem F., Lebrun M., Mestres C., Sieczkowski N., Boulanger R., Collignan A., “Transfer kinetics of labeled aroma compounds from liquid media into coffee beans during simulated wet processing conditions”, Food Chemistry 322:126779, 2020, DOI 10.1016/j.foodchem.2020.126779; Hadj Salem F., Achir N., Sieczkowski N., Boulanger R., Collignan A., “Modelling the transfer and degradation kinetics of aroma compounds from liquid media into coffee beans during simulated wet processing conditions”, Journal of Food Engineering 343:111303, 2023, DOI 10.1016/j.jfoodeng.2022.111303; Alliance for Coffee Excellence, results and auction of the Taza de Excelencia Colombia 2018, entry “El Paraiso, Diego Samuel Bermudez Tapia, Piendamo, Cauca, 89.76, Bourbón, Double Anaerobic”, sold at 54.10 dollars a pound; Europe PMC and Crossref queries, September 2026, on “thermal shock” and “choque térmico” associated with coffee, with no relevant result
    PreviousWet hulledNextCo-fermentation and infusion
    © 2026 Marco De Sousa Rosa · PXL RoastersTexts under CC BY-NC-SA 4.0 · tools: all rights reserved · images: see the provenanceHome · Colophon · Site map