Heat Transfer Fundamentals
Heat transfer is the practical link between profile settings and how coffee actually roasts. On Roest systems, the same temperature number can represent different heat delivery depending on airflow, agitation, batch size, pressure environment, sensor placement, and roast mode. This page explains those relationships and gives a compact adjustment map for changing heat transfer deliberately rather than chasing curve shape alone.
Temperature Is Not the Same as Heat
A temperature reading describes the probe environment; it does not by itself describe how much energy is reaching the coffee. Tom Roest summarized the distinction as “less heat, same temperature,” and noted that heat transfer is proportional to temperature difference along with other factors 2 sources. In practice, two roasts can show similar inlet or air temperatures while transferring different amounts of heat because the moving air, bean pile, and agitation are different.
Airflow changes the heat-transfer rate. Higher airflow at the same temperature transfers more energy, while lower airflow reduces heat transfer and requires higher temperature if the goal is to keep the same roast time 2 sources. Christopher Feran framed the same mechanism as airspeed affecting the heat-transfer coefficient, while the temperature delta remains the underlying driver source.
For profile work, this means inlet temperature should be read as one part of a heat-transfer system, not as a standalone “heat applied” number. For deeper treatment of inlet targets and offsets, see Inlet Temperature Management.
The Main Heat-Transfer Levers on Roest
Roest heat transfer is dominated operationally by hot air, but bean movement and batch fill determine how effectively that hot air contacts the coffee. Hot air enters the drum/chamber through inlet holes, and the internal paddles or propellers move the bean pile through that air stream 2 sources. This makes Roest more sensitive to airflow, RPM, and batch-size changes than a simple static-temperature model would suggest.
RPM is a major heat-transfer lever because it changes bean agitation and how often beans are exposed to the incoming hot air. Multiple discussions converge on the same practical rule: higher RPM increases heat transfer and tends to speed the roast, while lower RPM slows transfer 2 sources. Roest guidance quoted in the community also states that changing main motor speed has a more significant roast impact than airflow changes source. RPM-specific settings belong on Drum Speed / RPM Settings.
Airflow is the other primary lever. Higher airflow can increase transfer efficiency, but in temperature-based profiles it also uses more energy and can force the heater to work harder 2 sources. Lower airflow can require higher inlet temperatures for equivalent roast speed, and reducing both airflow and inlet together compounds the reduction in heat transfer, especially on P3000 2 sources. For practical fan settings, see Airflow and Fan Settings.
Batch size changes the air-to-bean ratio and therefore the entire heat-transfer behavior. A larger batch occupies more of the chamber, changes how the hot air travels through the bean pile, and can reduce how much heat escapes directly to the exhaust. This is why profile changes across 100g, 150g, and 200g are not simple linear power or inlet adjustments 2 sources. Batch strategy is covered in Batch Size Scaling.
Batch Fill, Air Escape, and Heat Retention
Small batches expose more empty chamber volume, so hot air can pass the beans and leave through the exhaust more easily. Larger batches can act more like a plug or filtration bed, making hot air travel more slowly through the coffee before exiting. Denis described this as a major reason larger Roest batches transfer heat better and lose less heat to the exhaust 2 sources.
This is also why the same inlet or ET number can behave differently by batch size. At the same 225–230°C stated range, Denis said most 100g batches may not crack, while 200g batches can crack at that range source. The point is not that 225–230°C is a universal crack target; it is that batch fill changes what the same temperature reading means in the roast.
Bigger is not automatically better for every goal, but heat-transfer behavior changes substantially once the roaster is loaded more fully. Several experienced users report that 150–180g or 180–200g batches behave very differently from 50–100g batches in heat retention, development time, and taste 2 sources.
Counterflow, Reversal, and Tilt
Counterflow and reversal-style operation improve heat-transfer efficiency by changing how beans meet the incoming hot air. Denis described counterflow as pushing heat better into the beans across batch sizes and said it can use inlet values roughly 40–60°C lower than normal-flow mode source. Tom Roest likewise framed counterflow as affecting heat-transfer efficiency, not simply how hard the heating element works to reach a set temperature source.
Tilting older S100-style machines can partially simulate this by lofting or positioning the bean pile differently. For one S100 approach, 185g with a 12–14° tilt was described as allowing inlet to be reduced by about 50°C, with the goal being lower inlet rather than copying Ultra values exactly source. Sorin also reported that tilting around 25° with 200g produced more than -30 Pa and allowed more than a 50°C inlet reduction source. curated If physically tilting a non-Ultra/S100-style machine, make sure the roaster, exhaust connection, bean cup, and chaff path remain stable and secure; do not improvise an unstable support or tilt in a way that could spill hot beans, kink exhaust ducting, or compromise smoke/chaff evacuation. For model-specific guidance, see Roest Ultra Guide and Pressure Management.
Pressure Versus Airflow
Roest’s manufacturer view is that heat transfer should be thought of mainly in terms of airflow, not pressure; Tom Roest stated that sample-roaster pressure differences are minuscule, that pressure value does not matter much in the system, and that negative exhaust pressure is primarily to ensure smoke exits through the exhaust 2 sources. Other experienced roasters have observed pressure-related roast differences and describe pressure environment as affecting transfer, density, expansion, or cup character 2 sources.
The practical synthesis is to avoid treating pressure as an independent flavor dial unless airflow, batch size, RPM, and inlet are also controlled. Pressure is useful diagnostically because it reflects system resistance and exhaust behavior, but heat-transfer changes should be interpreted through the actual roast outcome: time, crack behavior, exhaust behavior, color, weight loss, and cup. curated Do not reduce exhaust/negative pressure below what is needed for safe smoke and exhaust removal just to chase a pressure number or cup effect; coffee roasting exhaust can include smoke, CO, and combustible chaff, so ventilation and clear exhaust flow take priority over pressure experimentation. For pressure-specific calibration and troubleshooting, defer to Pressure Management.
Sensor Readings and Heat-Transfer Interpretation
Probe readings are not direct measurements of bean heat absorption. Probe placement, probe thickness, airflow, and batch size all affect what the curve shows. Denis noted that probe thickness changes delay and graph shape, and that Roest’s bean temperature sensor is inside above the hot-air inlet while the drum sensor is outside under the drum 2 sources. Patrick also warned that under high heat input, the BT probe can read too much hot air into the mix source.
Airflow does not necessarily change the steady-state inlet reading much, but it can change the lag and the amount of heat delivered at the same displayed temperature. Tom Roest stated that airflow has minimal impact on inlet reading accuracy at steady state, even though airflow affects heat transfer 2 sources. This is why curves should be evaluated together with cup results, color, weight loss, and physical bean inspection. For curve interpretation, see Bean Temperature Profiling and Rate of Rise Management.
Conduction, Convection, and the Roest Chamber
Roest discussions often use “conduction” in several ways: contact with metal surfaces, bean-to-bean contact, stored thermal energy in the chamber, and the smoothing effect of a larger bean mass. These should not be collapsed into one variable. The machine is commonly treated as primarily hot-air driven, with limited drum-mass contribution compared with a heavy cast drum roaster; however, early-roast contact, inlet plate effects, and bean-to-bean conduction can still matter 2 sources.
Some comments frame Roest as mostly hot air with little conduction from the drum, while others emphasize that conduction or stored-energy effects are present enough to affect roast behavior, especially early in the roast or at larger batch sizes. The safest operational language is to describe the controllable levers—airflow, inlet, RPM, batch fill, preheat state—rather than trying to assign a fixed percentage to conduction versus convection.
Environment and Moisture Effects
Ambient conditions affect heat transfer. Tom Roest stated that air humidity changes how much energy the air carries at a given temperature source. Christopher Feran also noted that less dense air at elevation makes convection less efficient source. This helps explain why a profile may need adjustment across locations, seasons, or ventilation setups even when the displayed machine settings are unchanged.
Moisture inside the bean also affects internal heat movement. Christopher described moisture as very effective at conducting heat relative to coffee structure, and Dave Ewald similarly stated that moisture helps conduct heat toward the center up to a point 2 sources. These claims support a cautious approach to very aggressive early heat: fast exterior drying can make the outside appear developed while the center lags. For phase timing and drying behavior, see Drying and Maillard Phases.
Heat Transfer and Defects
Many Roest defect discussions point to excessive hot air, excessive transfer rate, or poor match between surface heating and internal development. Tipping and scorching are often discussed as hot-air or fast-roast defects on air-driven systems, but reducing airflow is not a universal fix for every bean 2 sources. Bean inspection is important: dark spots near the tip or burned internal pores after cracking beans open were used as indicators of overly aggressive heat transfer 2 sources.
For symptom-level diagnosis and fixes, use Roast Defects Troubleshooting. This page should be used to understand why the fixes work.
Practical Heat-Transfer Adjustment Map
Use this section as the canonical quick-reference for changing heat transfer. Change one primary variable at a time when possible, and compare using the same coffee, batch size, charge/preheat state, and endpoint.
| Goal or symptom | Primary adjustment | Expected heat-transfer effect | Watch-outs |
|---|---|---|---|
| Roast is too slow, but inlet should not be raised | Increase RPM/agitation or airflow | More bean exposure to hot air; faster heat transfer and likely faster roast | Higher RPM can change readings and development; excessive agitation/air can increase surface stress |
| Roast is too fast or exterior looks heat-stressed | Reduce transfer rate by lowering airflow, lowering RPM, or lowering inlet depending on which lever caused the stress | Slower energy delivery to the bean surface | Lower airflow may require higher inlet to maintain the same roast time; do not assume lower fan always prevents tipping |
| Same roast time is desired with lower airflow | Raise air/inlet temperature | Compensates for lower transfer coefficient | Higher temperature can increase surface damage risk if held too long |
| Switching into counterflow or tilt-assisted roasting | Start with substantially lower inlet targets | Counterflow/tilt increases transfer efficiency; reported reductions are often around 40–60°C | Do not copy Ultra values one-to-one; tune to the machine and batch |
| Scaling from 100g to 150–200g | Rebuild the profile rather than linearly scaling power or inlet | Batch fill changes air-to-bean ratio, heat retention, exhaust behavior, and sensor readings | Larger batches may allow more late power reduction; smaller batches may need later or gentler power shaping |
| Lowering both fan and inlet at once | Avoid unless the goal is a large transfer reduction | Both changes reduce heat delivery | On P3000, reducing fan and inlet together was described as “doubling” the reduction of heat transfer source |
| Comparing pressure settings | Treat pressure as diagnostic unless airflow and resistance are controlled | Pressure reflects system behavior but is not a clean heat-transfer variable by itself | Negative pressure is also a smoke-management requirement, not just a roast-style setting |
| Curves look similar but cups differ | Evaluate batch size, airflow, RPM, and probe context before assuming equal heat transfer | Similar BT/IT curves can hide different delivered energy | Use color, weight loss, exhaust behavior, crack behavior, and cup data together |
Specific conversion examples should be treated as starting hypotheses, not universal settings. Counterflow conversions have been described as needing roughly 40–60°C lower inlet than normal flow, and a 185g S100 roast with 12–14° tilt was described as allowing about a 50°C inlet reduction 2 sources. For batch scaling, the recurring rule is not “more beans equals more power” or “less beans equals less inlet,” but “different batch size equals different heat-transfer system.”