A General Design for Energy Test Procedures

Appliances are increasingly controlled by microprocessors. Unfortunately, energy test procedures have not been modified to capture the positive and negative contributions of the microprocessor to the appliance’s energy use. A new test procedure is described, which captures both the mechanical and logical features present in many new appliances. We developed an energy test procedure for refrigerators that incorporates most aspects of the proposed new approach. Some of the strengths and weaknesses of the new test are described. WHY ARE NEW ENERGY TEST PROCEDURES NEEDED? The technologies employed in major appliances are undergoing a major transformation. In the past, most aspects of an appliance’s operation were controlled by the user. New appliances, however, have microprocessor controls, which may adjust the appliance’s operation without any action by (or even knowledge of) the user. The microprocessor can gather information through sensors or from memory of previous cycles to select an operating strategy that results in enhanced amenities or services to the user (Meier 1997; Meier 1998). This trend has the potential to save energy, water, and other resources in many different ways. For example in washing machines, sensors may measure the weight of clothing and the extent of soiling. The microprocessor will use this information to select the minimum amount of water and detergent to achieve clean clothes. Microprocessors can also control variable-speed motors in air conditioners and refrigerators; this will allow better temperature regulation, dehumidification, often with less energy than traditional approaches. Unfortunately the energy savings from microprocessor controls are not fully captured in the present energy test procedures. Some omissions in the present tests are that they: • Fail to include part-load conditions (less than full loads in washing machines, cooling or heating at less than steady-state output, etc.) • Ignore learning capability from previous cycles • Ignore sensing special conditions for service (such as level of soiling and type of fabric) • Fail to recognize communication between the appliance and a network (including the internet) In some cases these omissions lead to only a small discrepancy between the laboratory measurements and actual use, but this discrepancy is likely to grow as microprocessors become more sophisticated. At the same time, some manufacturers are programming the microprocessors in appliances to recognize when the appliances are being tested. When the unique test conditions are identified, the microprocessor modifies performance such that it uses less energy than it would under ordinary conditions. This paper outlines a general approach to energy tests for this new generation of appliances. These tests aim to assess the integrated performance of both the appliance’s mechanical features (the “hardware”) and its logical features (the “software”). ELEMENTS OF THE NEW APPROACH In the new approach, the test procedure consists of “hardware tests” and “software tests.” The results of these tests are inputs to a model that simulates the appliance’s energy use. The simulation model programmed to predict energy use in any conditions requested by the user.