Air-source heat pumps have one weakness: they get less efficient exactly when you need them most. On the coldest night, the outside air — the very thing the pump is trying to extract heat from — is frigid, and the pump strains. A 2021 patent leans on a simple geological fact to dodge that problem, and then adds a second trick on top: a tank that lets the system do its hardest work when nobody is watching.

The fact: a few feet below the surface, the ground holds a remarkably steady temperature all year, neither summer-hot nor winter-cold. A ground-source heat pump exchanges heat with that stable reservoir instead of the swinging outdoor air. In winter it pulls heat from ground that is still moderate; in summer it dumps heat into ground that is cooler than the air. The reservoir never goes to the extremes the pump dreads.

“An intelligent heat pump system having a dual heat exchanger structure includes a heat source side heat exchange member, a heat pump, an external expansion valve, a refrigerant-water heat exchanger, a heat storage tank, and a target side end unit.”— U.S. Patent No. 11,092,360 source

Grant US11092360B2, “Intelligent heat pump system having dual heat exchanger structure,” takes the steady-ground idea and engineers around two heat exchangers instead of one. On the ground side, a “heat source side heat exchange member” trades heat with the earth. On the other side, a “refrigerant-water heat exchanger” feeds a “heat storage tank” of water. The clever part is what the system does when no one is calling for heating or cooling.

Here is the mechanism the document describes. The system runs in two distinct modes. In the “non-air-conditioning state for the target site,” the patent explains, “while the refrigerant is circulating between the heat pump and the refrigerant-water heat exchanger, the water is cooled or heated, and the cooled or heated water is stored in the heat storage tank.” In other words, during idle hours the machine is not off — it is quietly charging a thermal battery, chilling or warming a tank of water. Then, when the building does need conditioning, “the heat pump and the heat storage tank supply cooling and heating to the target site,” so the stored water shoulders part of the load and the compressor does not have to do it all in real time.

Why split the work this way? The patent gives two reasons, and both are about efficiency and timing. First, the dual exchanger lets the refrigerant cycle run gentler: “it is possible to increase the coefficient of performance of the intelligent heat pump system by lowering the condensation temperature during cooling of the heat pump and raising the evaporation temperature during heating.” The coefficient of performance is simply how many units of heat the pump moves per unit of electricity it burns, and squeezing the temperature gap the refrigerant has to bridge is the most direct way to raise it.

Second is the demand-shifting argument, which is where the storage tank earns its keep. The system can “realize summer cold storage or winter heat storage during an extra non-air-conditioned time period such as nighttime, non-operation time, and so on, thereby reducing the power consumption… in the normal operation and enabling the demand management of the building or the power network through the reduction in the peak power, and preventing the black-out due to the power peak.” Charge the tank overnight on cheap, abundant power; draw it down during the afternoon peak. The heat pump becomes a tool for grid demand management, not just comfort.

The dual design also quietly cuts the most expensive part of a ground-source install. The patent notes that when ground heat transfer is weak — “when the geothermal heat conductivity or the like is low so that the amount of heat exchange in the ground becomes insufficient” — the stored cold or warm water “can assist such situations. Thus, the number of the geothermal drilling holes or the size of the heat source side heat exchange member can be greatly reduced, thereby reducing the installation and operation costs.” The thermal tank is a substitute for boreholes you would otherwise have to drill, which directly attacks ground-source's biggest barrier.

The two operating modes are worth seeing side by side, because they show how the tank changes the machine's job. In a “cold storage operation,” run during idle hours, the refrigerant loses heat in the heat pump, expands through the external valve, then absorbs heat from the tank water in the refrigerant-water heat exchanger — chilling that water and storing it for later. In a “cold air supply operation,” run when the building actually needs cooling, the patent describes the refrigerant being “primarily-condensed” in the heat pump and then “secondarily condensed through heat exchange with the cooled water supplied from the heat storage tank.” That second, pre-chilled condensing stage is the efficiency payoff: the refrigerant gives up heat first to the ground loop and then again to water that was cooled overnight, so the compressor faces an easier job at the moment of peak demand. The same structure runs in reverse for heat storage in winter. It is a heat pump that pre-positions its own cold and warmth, and the dual exchanger is what makes that staging possible.

Because that is the real story. Ground-source systems require digging — boreholes or trenches — which makes the upfront cost and site requirements much steeper than hanging an air-source unit on a wall. The efficiency is better; the installation friction is worse. The 2021 grant is worth reading as a marker of where the engineering attention sat: not on inventing the geothermal idea, which is old, but on adding a thermal-storage tank and dual-exchanger control to wring more performance from a steadier reservoir while shrinking the drilling that makes geothermal expensive. The earth's steady temperature is free; the patents are about using it well — and about doing the work at the right hour.