Interlaboratory verification of feed-forward compensation for enhanced reproducibility of emulator-type load-based tests – Case study of defrost operation –
Frost formation is among the main causes of performance degradation in heat pumps. As defrosting cycles are energy intensive and dynamic maneuvers programmed within the system control board, their characterization intensifies the challenges in reproducibility of dynamic performance tests and remains mostly unresolved. Emulator-type load-based tests respond to the necessity of performing representative yet reproducible performance evaluation of air conditioners and heat pumps. Core of this testing methodology is the use of a room emulator to mimic the response of a virtual building to the capacity supplied by the tested unit under given load conditions. Thereby allowing for tests conducted while operating the system with its native control and while defining representative thermal characteristics of the conditioned space, independently from the thermal and moisture inertia of the testing facility itself. The feed-forward compensation method (FFC) is a model predictive control technique developed to minimize information loss and discrepancies in the continuous data transfer between hardware and software side of such testing methodology. This study practically deploys the FFC in a feed-forward architecture during actual tests on a 10kW-ceiling-type unit operated in heating mode and exhibiting defrost dynamics. Maximum peak to peak delay during cycling on-off operation was limited to below 60 seconds. The benefit on results reproducibility was investigated and verified through interlaboratory tests conducted in two different facilities. Consequently, reached a level equivalent to current fixed-capacity tests, with maximum COP deviations of the same tested unit within 2% in the tested range of conditions. The proposed method does not require additional equipment and aims at enhancing the capability of current testing equipment in conducting reliable load-based and dynamic tests, opens to reliable and clear investigations of defrosting operation, and complementarily enables supports control optimization to minimize performance degradation due to frost formation.