Customers have complained of hydraulic hammering noises on heat pump systems for years. Field people refer to it as the “knock-knock” noise. The complaints usually surface at the end of the heating season. The noise apparently occurs following a heating cycle during the light load part of the year. This means it could occur sometime in March or April and reportedly is most noticeable in the early hours of the morning. The compressor cycles off after a heating cycle and the hydraulic hammering or “knock-knock” noise commence shortly thereafter.
Theory - cause of the hydraulic hammering noise
Hydraulic hammering can occur with abrupt impedance (a valve closes) to non-compressible fluid flow (liquid). This can create an initial shock wave which can then resonate through the system piping. During the heating mode, the internal check valve in a thermostatic expansion valve (TEV) would normally be in the open position on the indoor coil with reverse flow through the valve. The check valve on the indoor coil’s TEV would tend to close after the compressor cycles to the off position. However, gravity would tend to hold the check valve in the open position if the TEV was in a vertical position with the thermostatic element assembly (also known as the power head in general industry vernacular) pointing down towards the earth even after the compressor cycles to the off position. This could promote normal forward flow through the indoor coil TEV during the system pressure equalization process. Forward flow through the check assembly does cause the check valve to close by design. This could account for the necessary abrupt change that must take place to create the initial shock wave thus creating the hydraulic hammering phenomenon. If this theory has any merit, repositioning the TEV to eliminate the force of gravity would reduce or curtail the hydraulic hammering problem.
Refer to Figure 1; the Type CBBI valve is depicted with the check valve in both a Closed and an Open position.
Site & system details
Previous field trips to supposed trouble job locations yielded negative results, meaning no objectionable noise was ever observed, much less repeated. That was about to change with this trip. The trouble job was located in a historic home (1870 vintage) in South Central Texas that had been restored and remodeled with the input of an architect. It was fitted with acid washed concrete floors, stone walls, hardwood accents and a relatively new heat pump system supplied by one of the major equipment manufacturers.
The heat pump system utilized R-22 as the refrigerant and was fitted with the CBIVE-2-GA on both the indoor and outdoor coils. The air handling unit and indoor coil were mounted in the ceiling in close proximity to the bedroom; the equipment was relatively new and had been supplied and installed via reputable means. If this system exhibited the so-called “knock- knock” noise, it would certainly reverberate through-out this structure of hard surfaces.
A scroll type compressor with a discharge line check valve was deployed on this unit; the discharge line check valve is intended to prevent the compressor from running in a reverse direction following an off cycle. The system was originally installed with the TEV in a horizontal position in the indoor coil. The line set was within the OEM’s specified requirements regarding size and design. The piping was routed through the attic and above grade.
The unit had been previously re-charged with R-22 to confirm system integrity and removal of non-condensable contaminants. The system was checked for proper superheat and sub-cooling performance while we were at the site. Superheat was approximately 12°F and sub-cooling was approximately 11°F. These numbers were all within OEM specifications. The system heated and cooled adequately and there was never a complaint regarding this aspect of performance.
Based on the theory, we decided to first remove the Chatleff style piston assembly in the distributor. We speculated that perhaps the piston which serves as both a check valve for reverse flow and a nozzle assembly for forwarding flow was contributing to the hydraulic hammer perhaps in conjunction with the TEV’s check valve. The system was pumped down and the piston assembly was removed. The system charge was not modified.
Even with the piston assembly removed from the distributor, the hydraulic hammering noise could reliably be produced; however, the time duration of the noise was shortened considerably. We reinstalled the piston assembly in the Chatleff distributor.
We then repositioned the indoor coil TEV so as to be totally upright; i.e., in a vertical position with the thermostatic element pointing towards the sky. We were able to accomplish this by simply rotating the TEV as enough “slack” existed in the system piping. Again, this was performed without disturbing the system charge in any way. This solved the problem on this trouble job. Merely rotating the TEV to prevent gravity from opening the check valve following the heating mode cycle prevents the hydraulic hammering noise. We then attempted to reproduce the noise over the course of two days and were unable to reproduce it while prior to repositioning the TEV it could easily be done.
The problem occurs only when the necessary system conditions exist. It appears the indoor coil TEV needs to be upside down or at least on its side and TEVs must be present on both the indoor and outdoor coils for the noise to occur. It may also be necessary for the discharge check valve on the scroll compressor to have some associated leak rate; it is speculated this may contribute to the necessary system conditions for noise to occur.
Refer to Figure 2 for recommended TEV positions when equipped with the internal check valve. These recommendations pertain to TEVs equipped with internal check valves that are installed in the valve body so as to be parallel to a vertical line through the center of the TEV. The Sporlan Type CBI and CBBI thermostatic expansion valves are examples of this type of construction.
The hydraulic hammering noise would not surface with any other expansion device on the outdoor coil; both the indoor and outdoor coils must be equipped with a thermostatic expansion valve. If the outdoor unit was fitted with a fixed tube orifice of any kind or a TEV with a bleed port, the noise didn’t occur. Apparently, this provides a “vent “or release for the pressure differential after the compressor cycles to the off position and the shock wave never occurs. In all cases, the problem can occur with conventional and balanced ports alike and includes most any competitive product that utilizes a gravity influenced check valve for reverse flow for a bypass around the main port.
Furthermore, the noise issue is not a valve malfunction. The thermostatic expansion valve controls superheat at the bulb location in spite of the noise issue. The problem only occurs when the necessary system conditions exist. It appears the indoor coil TEV needs to be upside down or at least on its side and non-bleed style TEVs must be present on both the indoor and outdoor coils for the noise to occur. It may also be necessary for the discharge check valve on the scroll compressor to have some associated leak rate.
And finally, with the TEV in a vertical position and the thermostatic element pointing towards the sky, the noise will not occur.
For more information about TEVs see Bulletin 10-10 Thermostatic Expansion Valves.
Article contributed by Jim Jansen, senior application engineer, Sporlan Division of Parker Hannifin
Additional resources on HVACR Tech Tips: