Foster-Miller developed a "multiplexed" refrigeration system to save electrical energy. The multiplexed configuration uses one centralized compressor rack to hold a few efficient compressors and services every cooled display case and storage freezer in a supermarket. Proper rack design permits servicing five or more distinct refrigeration temperature ranges. Proper control offers excellent efficiency for low to high cooling capacities with minimal compressor cycling. Our refrigeration system demonstrated a 25 percent energy savings over conventional systems, despite the prohibition against using the older, more efficient ozone-depleting refrigerants. This system was the ultimate result of a broad effort that included a detailed system modeling exercise in combination with the design, construction, and control of an optimized refrigeration system.

The first step in this program was the creation of a detailed thermal model of the target supermarket. The objective was to model the hour-by-hour loads on the supermarket refrigeration system and determine the annual operating costs of equipment configurations in response to those loads.

The model evaluated the effect of air flow throughout the store (via infiltration and door-openings), the statistical frequency of customers opening display cases, compressor performance based on manufacturer's detailed data, and temperature and relative humidity changes monitored hourly and contained in historical meteorological data. Taking all these effects into account, the supermarket simulation accurately predicted both the cost of power (a utility's charge for kilowatt demand) and the total kilowatt-hours consumed on a daily, monthly, and yearly basis. 

To validate the model, an elaborate monitoring system was created and installed in the target supermarket.  A key aspect of this system was remote telemetry. From our offices in Massachusetts, we were able to monitor the detailed real-time behavior of the target supermarket in California. The data was then compared to the output of the supermarket thermal model to refine and validate the model.

The supermarket model was then used to investigate various "what if" scenarios. The model enabled many different candidate cooling system configurations to be examined at low cost and in a short amount of time without actually constructing the systems. For example, regarding simple compressor efficacy, we compared liquid injection, variable speed, and internal compounding (a style of two-stage compressor) for off-the-shelf models of the two leading compressor manufacturers.

The result of this work was an optimized cooling system that included three to six unequally sized, parallel compressors (rather that a collection of small, independent and less efficient compressors on each display case) and a microprocessor-based controller. Novel energy saving features such as floating condenser pressure, mechanical sub-cooling, external liquid suction heat exchangers and demand frost sensors were incorporated into the design.

Return to Human Factors

Printable Version