Heat pump makers: controllers tuned in the lab, not the field

Controllers are tuned on the lab unit, not per installation

What happens. A heat pump controller is tuned on a test unit in the laboratory. In the field, every house has its own pipes, flow temperatures, hot-water habits and buffer tank, and the same settings fit none of them well.

What it costs. The result is lower seasonal efficiency than on the test bench, short cycling, noise and complaints from customers. The service technician adjusts what he can and leaves.

Why the usual fix fails. Wider tolerances in the controller avoid faults but do not give good efficiency. A tuning guide for installers is followed in some places and not in others.

How a fixed-price remote diagnosis checks it. From field data of installed units (temperatures, compressor speed, power, set curves), I check how far the real operation differs from the lab case. The fixed-price remote diagnosis tells you whether a self-adapting controller can learn the installation and where it would gain most.

Grid operators may dim heat pumps and chargers

What happens. Under the rule that lets grid operators temporarily reduce the power of heat pumps and wallboxes when the local grid is overloaded, a heat pump must accept a dimming window (Federal Network Agency). In that window the compressor may run only at reduced power.

What it costs. The house gets cold, hot water runs short and the customer complains, or the guarantee for comfort is at risk. A heat pump that simply stops loses its efficiency on restart.

Why the usual fix fails. A bigger buffer tank is costly and takes space. A fixed rule that heats more before the usual hours does not know when the next dimming comes.

How a fixed-price remote diagnosis checks it. From operating logs and the building’s thermal response, I check how far ahead a controller has to heat to ride through a window. The fixed-price remote diagnosis tells you whether predictive control using the dimming signal would keep comfort and what it would need in the unit.

Industrial heat pumps get bigger, with multi-stage compressors

What happens. Heat pumps for industry grow to megawatt units with several compressor stages, large heat exchangers and new refrigerants. They are no longer scaled-up home units.

What it costs. Operation is unstable in transients, such as when the heat source or the demand changes quickly. This costs efficiency, and units trip rather than adjust.

Why the usual fix fails. The controllers of smaller units do not scale: interactions between stages and the slow heat exchangers are new. Commissioning by trial is slow and expensive for a unit this size.

How a fixed-price remote diagnosis checks it. From commissioning logs of existing large units and the process diagrams, I check where the instabilities come from. The fixed-price remote diagnosis tells you whether a model-based control of the stages and the refrigerant circuit would stabilise operation, and what model it needs.

Defrost runs on timers

What happens. The outdoor unit ices up in cold, humid weather, and the heat pump defrosts it. Many defrost on a timer or on a temperature difference that is a crude proxy for the ice.

What it costs. Defrosting too often wastes energy and heats the room less. Defrosting too late lets the evaporator ice over and the efficiency collapse.

Why the usual fix fails. A different timer setting suits one site and not the next. Adding a sensor for ice is another part that fails in frost.

How a fixed-price remote diagnosis checks it. From unit logs (air and evaporator temperature, fan speed, compressor power, defrost events), I check how well the frost on the evaporator can be estimated. The fixed-price remote diagnosis tells you whether defrost on demand would save energy and how it would trigger.

A new propane refrigerant brings new dynamics

What happens. Heat pumps are moving to propane as refrigerant. Charge limits are small, expansion valves and compressors behave differently, and the controller tuned for the old refrigerant does not carry over.

What it costs. Units show instabilities in the new refrigerant that were not seen before, and each new model needs a new tuning round. Delays hit launches.

Why the usual fix fails. Copying the old parameters gives a unit that runs, but not efficiently. Overdesigning the controller to be safe costs efficiency.

How a fixed-price remote diagnosis checks it. From test records of units with the new refrigerant, I check how the dynamics differ and which controller parts need to change. The fixed-price remote diagnosis tells you whether a control adapted to the refrigerant, with a model rather than trial, would cut the tuning effort.

Modulating compressors cycle at part load

What happens. Variable-speed compressors can run at low power, but below a threshold they switch on and off instead. The heat pump short-cycles in mild weather, when the demand is smallest.

What it costs. Short cycling wears the compressor, wastes energy at each start and brings noise. The customer’s heating gets less even.

Why the usual fix fails. A larger buffer tank reduces cycling but costs space and heat. A tighter hysteresis setting gives a poor temperature curve and complaints.

How a fixed-price remote diagnosis checks it. From unit logs at part load (compressor speed, flow temperature, start counts), I check how the capacity control behaves near the lowest speed. The fixed-price remote diagnosis tells you whether a model-based capacity controller could keep the unit running smoothly at low demand.

Lab ratings are not reached in the field

What happens. The data sheet shows the seasonal efficiency measured in the laboratory. In the customer’s house the unit does worse, sometimes much worse, and the installer and the maker blame each other.

What it costs. Complaints, service visits and goodwill gestures cost money, and the reputation of the whole product line suffers.

Why the usual fix fails. Checking every installation on site is not possible. Remote data from the unit show symptoms but not the cause: pipe sizing, flow temperature, setting or fault.

How a fixed-price remote diagnosis checks it. From remote data of installed units, I look for patterns that separate installation faults from product faults. The fixed-price remote diagnosis tells you whether a field-data diagnosis can pick out the units that are not performing and say why.

Every product variant is tuned by hand

What happens. A heat pump family has many variants: sizes, refrigerants, heat sources, hot-water options. Each is tuned on a test stand by an engineer, and the settings are copied from a neighbour as a starting point.

What it costs. Engineering time per variant is high and the number of variants keeps rising. Variants end up with different quality of tuning, depending on who did it.

Why the usual fix fails. A tuning guide that says ‘copy from the next size’ ignores that the dynamics change with the size. More test stands only move the cost.

How a fixed-price remote diagnosis checks it. From the tuning records and test data of existing variants, I check how much of the tuning follows from the unit’s physical data. The fixed-price remote diagnosis tells you whether a parameter set per variant can be derived from a model, and which variants still need a real test.

Models and control for hybrid heat pumps and home energy need specialists

What happens. Hybrid heat pumps combine a heat pump with a boiler and often a solar system, a battery and a charger. The control has to decide, every few minutes, which source covers which demand.

What it costs. Teams that develop these systems find it hard to fill model-based control roles. The result is a simple rule set that works, but not at the lowest cost to the household.

Why the usual fix fails. A rule set grows complicated with each new component, and nobody can say how far it is from the best. Tuning the rules on a test house does not carry to other houses.

How a fixed-price remote diagnosis checks it. I review the existing rules together with logs from a few houses and compare them with a model-based schedule. The fixed-price remote diagnosis tells you what the difference is, and whether a predictive controller is worth building in your product.

High-temperature industrial heat pumps are not yet routine

What happens. Industry wants heat pumps that deliver steam-grade temperatures from waste heat. Large units at these temperatures are not yet fully proven, and few plants have experience integrating them into a running process.

What it costs. Without that know-how, integration is risky: the heat pump must follow a process that changes while the process cannot be allowed to suffer. Projects are put off for fear of surprises.

Why the usual fix fails. A static design calculation shows the average, not the dynamics of the process around it. Putting in a storage tank adds cost without a plan for how to run it.

How a fixed-price remote diagnosis checks it. From the process data, heat demand profiles and the planned heat pump, I build up what a model of the integrated system needs. The fixed-price remote diagnosis tells you whether the control of heat pump and storage can be tested before the investment, and which operating strategy to try first.