Skip to content
Website
🧱 CFC 241Commissioning, control and acceptanceChapter 1 of 10 in the section

Setting the heating curve: slope, parallel shift and correction by occupant feedback

This chapter details the method for adjusting a weather-compensated control according to the outdoor temperature without destabilising the heating system. You will learn to separate the mathematical effect of the slope from that of the parallel shift. You will be able to correct comfort drift in the mid-season and in severe cold using a methodical measurement protocol proven on Swiss construction sites.

Key takeaways

  • The slope changes the inclination of the curve and acts mainly at very low outdoor temperatures.
  • The parallel shift moves the entire line and corrects the temperature deviations observed in the mid-season.
  • Any adjustment on a controller requires a minimum observation period of twenty-four to forty-eight hours.
  • The SIA 384/1 standard requires a thermal design that guarantees indoor comfort without excessive flow temperature.

Separating the effect of the slope and of the parallel shift on the flow water

On the heating controller, the heating curve links the outdoor temperature measured by the facade sensor to the flow temperature required at the emitters. Far too often on the construction site, technicians change either the inclination or the shift indiscriminately as soon as an occupant reports discomfort. This confusion ruins the balancing of the system and degrades the overall efficiency of the generator, particularly when a heat pump is operating. The slope, also called the steepness or multiplying coefficient, determines the variation of the flow temperature for each degree lost outdoors. The greater the heat losses of the building envelope, the steeper this slope must be to compensate for the heat losses calculated according to SIA 384/1.

Comparison of weather-compensation parameters
ParameterCommon namePhysical effectMain period of influence
Slope (S)Curve inclinationMultiplies the difference between the room setpoint and the outdoor temperatureSevere cold (below 0 °C)
Parallel (K)Shift or translationAdds or subtracts a constant value in Kelvins along the entire lineMid-season (between 5 °C and 15 °C)

The parallel shift moves the entire curve up or down by a fixed value expressed in Kelvins. Unlike the slope, whose effect is zero at the neutral point of the room setpoint, the shift acts uniformly under all weather conditions. On mild autumn or spring days, the difference between outdoors and the indoor setpoint remains small. At that point, the mathematical influence of the slope on the calculation of the flow temperature becomes minimal. It is therefore the parallel shift that governs heat output in the mid-season. If the slope is increased in mild weather to overcome a feeling of cold, massive overheating results as soon as winter frost arrives.

The SIA 384/1 standard sets the rules for calculating basic heat losses and serves as the reference for determining the nominal flow temperatures according to the chosen emitters, underfloor heating or radiators.

norme SIA 384/1

For systems fitted with underfloor heating, the slope typically ranges between 0.3 and 0.8 in French-speaking Switzerland, depending on the insulation of the envelope. For low-temperature radiators, the values lie between 1.0 and 1.5. Changing the slope rotates the line around its origin, fixed at the desired room temperature. Trying to make up a room temperature shortfall at 10 °C outdoors by raising the slope to 1.8 on a radiator circuit causes the flow to overheat in severe cold, leading to short cycling or a drop in the efficiency of the machine.

Correcting the curve by successive measurements without destabilising the system

To calibrate a heating curve without distorting the input data, the terminal control loops must be neutralised. During adjustment campaigns, all thermostatic valves and manifold actuators in the north-facing reference rooms, free of internal heat gains, must be 100 % open. The outdoor sensor must be installed on the coldest facade, protected from solar radiation and from ventilation heat discharge. Any thermal diagnosis carried out in the presence of direct solar gains or fireplace fires inevitably leads to a wrong parameter choice.

Decision matrix for adjusting the heating curve
Finding in the reference roomMild outdoor conditions (> 5 °C)Very cold outdoor conditions (< 0 °C)
Room temperature too lowIncrease the parallel shift by +1 K to +2 KIncrease the slope by 0.1 to 0.2
Room temperature too highDecrease the parallel shift by -1 K to -2 KDecrease the slope by 0.1 to 0.2
  1. Neutralising the terminal devicesFully open the emitter valves of the reference room and disable the influence of any room sensor in order to isolate the strict effect of the outdoor curve.
  2. Recording the initial situationRecord the stabilised room temperature with a reference thermometer, the outdoor temperature, the calculated setpoint and the actual flow temperature on the secondary circuit.
  3. Adjusting a single parameterChange only the slope or the parallel shift depending on the weather season, in gradual steps never exceeding 0.1 on the slope or 2 K on the shift.
  4. Thermal waiting periodAllow the building to absorb the change for a minimum of twenty-four hours for a radiator circuit, and forty-eight hours for a heavy underfloor heating system, before taking any new reading.

Patience remains the heating engineer's main tool when adjusting a curve. User complaints often come from interventions made too close together, where the technician compensates for the lack of inertia with a sudden rise in the setpoint. By proceeding in small increments and respecting the building's reaction times, comfort is ensured without consuming unnecessary energy.

Intervention matrix: separating the action in severe cold and in the mid-season

The frequent error in weather-compensated control lies in confusing the pivot point of the heating curve. The slope of the heating curve acts like a lever whose fixed point usually corresponds to the room setpoint of 20 °C. The further the outdoor temperature drops below 0 °C, the longer this lever arm becomes and the stronger the impact of a slope change on the flow temperature. Conversely, in the mid-season between 8 °C and 14 °C, the distance from the pivot point is small: a slope change has only a negligible effect on the fluid temperature. This is where the parallel shift, or translation, comes in. It raises or lowers the heating curve by the same value at every point of the curve, producing a constant thermal effect at -10 °C as well as at +12 °C.

Decision matrix for setting the heating curve
Outdoor weatherIndoor findingSlope actionParallel action
Severe cold (T out < 0 °C)Continuously too coldIncrease by +0.1 (+0.05 for underfloor heating)Do not change
Severe cold (T out < 0 °C)Continuously too warmDecrease by -0.1 (-0.05 for underfloor heating)Do not change
Mid-season (8 °C < T out < 15 °C)Room too coldDo not changeIncrease by +1 K to +2 K
Mid-season (8 °C < T out < 15 °C)Room too warmDo not changeDecrease by -1 K to -2 K

Where discomfort is present both in mild weather and in severe cold, changing the slope and the translation together destabilises the system. Touching both variables at once makes it impossible to identify the determining factor and distorts the analysis of the thermal behaviour. If the dwelling lacks heat at 10 °C as well as at -5 °C, only the parallel shift should be corrected at first. This general translation raises the water temperature in the mid-season without altering the shape of the heating line. Only after a stabilisation period will the behaviour in severe cold show whether the slope should be steepened or flattened.

On-site calibration protocol and compliance with the criteria of SIA 384/1

The choice of the initial slope is based on the sizing of the emitters and on the heat losses calculated according to SIA 384/1. This standard sets the base outdoor temperature according to the weather stations and climatic zones of Switzerland, from -8 °C on the Lake Geneva plateau to -14 °C or -16 °C in the Pre-Alps and the Jura massif. The inertia class of the envelope directly affects the reaction time of the system: massive concrete masonry smooths out daily fluctuations, whereas a lightweight timber-frame construction reacts immediately to changes in the outdoor air. The correction method by successive measurements requires the perceived thermal comfort to be objectified through repeated instrumented measurements.

  1. Neutralising the limiting effectOpen the manual valves fully and remove the thermostatic heads or actuators in the pilot rooms. If these devices remain active, they restrict the water flow as soon as the room overheats and mask a curve set too high.
  2. Recording the actual temperaturesMeasure the room temperature with a precision thermometer placed in the centre of the room at 1.50 m above the floor, out of direct radiation. At the same time, record the outdoor value measured by the facade sensor and the actual flow temperature.
  3. Adjustment in minimal incrementsApply a very gradual correction step to the control: 0.05 to 0.1 point on the slope for underfloor heating, 0.1 to 0.2 point for radiators, or 1 K to 2 K on the translation. Record the change in the boiler room logbook.
  4. Delayed observation of the behaviourWait until the whole building reaches its new thermal equilibrium before carrying out a second check. With metal radiators, equilibrium is established in 12 to 24 hours. With underfloor heating, the inertia of the screed requires the system to be observed for a full 24 to 48 hours.

According to art. 364 CO, the contractor is liable for the proper execution of the work in accordance with the recognised rules of the trade. This professional duty includes the methodical setting of the control, which is essential to reach the energy performance guaranteed at handover. Correcting the curve without haste protects the machine against premature wear and ensures lasting comfort without waste.

What to do in case of persistent drift or failed curve calibration

When successive adjustments of the slope and the parallel shift do not stabilise the indoor temperature, the controller is often compensating for a physical fault in the system. A heating curve never solves a distribution problem. Before changing the weather-compensation parameters again, the heating engineer checks the integrity of the measuring loop and the operation of the hydraulic components in the boiler room.

  • Checking the position of the outdoor sensor: check for the absence of stray direct sunshine, of the influence of a light well or of an extract-air ventilation outlet.
  • Checking the mixing valve: make sure of the direction of rotation of the actuator, the absence of internal mechanical seizing and the consistency between the control signal and the actual position of the plug.
  • Measuring the circulating flow: inspect the strainer filter on the heating return for fouling and check the differential pressure of the main circulator.
  • Recording the actual temperatures: compare the setpoint calculated by the control with the value measured by a surface-mounted thermometer on the direct or mixed flow.

If the building shows general overheating in the mid-season despite a lowering of the parallel shift, or persistent cold during winter peaks, the installer resets the values according to the initial calculations of SIA 384/1. He neutralises the influence of the room temperature, fully opens the thermostatic valve bodies of the reference rooms, then applies the correction method by successive measurements, observing a thermal stabilisation period of 24 hours for radiators and 48 hours for underfloor heating. Any premature intermediate adjustment prevents the building from stabilising and distorts the diagnosis.

Frequently asked questions

Why does the indoor temperature drop in the mid-season when it is correct in severe cold?

This symptom indicates an insufficient parallel shift masked by a heating curve slope that is too steep. In severe cold, the slope makes up for the initial shortfall by raising the flow temperature strongly. But when the outdoor temperature fluctuates between 8 °C and 14 °C, the flow becomes too low. The parallel shift must be raised by two to three degrees and the slope of the heating curve lowered at the same time.

What is the consequence of a slope that is too steep and restricted by thermostatic heads?

An excessive slope makes the water rise to too high a temperature. The thermostatic heads close to limit the overheating, drastically reducing the flow rate of the system. On a heat pump, this situation leads to short cycling of the compressor, a deterioration of the seasonal efficiency, premature wear of the components and parasitic hydraulic noise caused by the rise in differential pressure across the closed valves.

How long should you wait between two changes to the weather-compensated control?

The inertia of the building requires a strict waiting period before any new measurement. For a network of metal radiators, a delay of 24 hours without solar gain is mandatory. For distribution by underfloor heating embedded in a cement screed, 48 hours must be allowed for the room temperature to stabilise completely before making a further correction to the heating curve.

How do you tell a curve fault from a hydraulic balancing problem?

A heating curve fault affects all rooms uniformly: the whole building is too warm or too cold at the same time. Conversely, if some rooms stay cold while others reach their setpoint with correct flow temperatures, the problem lies in the distribution network. It then requires adjustment of the balancing valves and flow rates rather than a change to the weather-compensation settings.

Texts and standards cited

  • SIA 384/1Heating systems in buildings - Heat requirements and design capacity

In the same section