Refrigerated warehouses in the Northwest have significant demand response potential, due to their relatively large loads and, in some cases, flexibility of load provided by large thermal mass and acceptable range of product temperatures. In June 2019, the RTF reviewed and approved an analysis of this potential, including the stock characteristics, demand response opportunities, demand response mechanisms, savings potential (as a percentage of site load), and implementation costs.
Refrigerated Warehouses DR Workbook RTF Presentation DR Subcommittee Presentation DR Subcommittee Notes
Demand Response Site Potential
Estimates of average demand response potential per site were developed, based on a literature review of case studies and discussion with subject matter experts. The estimates were validated by subject matter experts and the RTF Demand Response Subcommittee. Note that these are very rough estimates of average potential across sites; the case studies illustrated a very wide range of achievements across sites. Some sites would be excluded from this potential because of constraints on product temperatures.
The following demand response potentials were estimated:
- 20-30% of site loads using existing equipment
- 40% of site loads with improved controls, part load capabilities, and product temperature instrumentation
The following event durations would be possible, assuming sufficient advance notice of an event (perhaps 12 hours) to pre-cool:
- Freezers: 6-8 hours
- Coolers: 2-4 hours
OPPORTUNITIES
The primary demand response opportunity at refrigerated warehouses is shifting the timing of refrigeration load. Often, this is achieved through pre-cooling before an event and then reducing refrigeration load during an event. For example, freezer spaces can typically lower air temperatures by 10⁰F in 10 hours prior to an event and then curtail load for 6-8 hours as temperatures rise back to the intended setpoint.
Freezers generally have the best potential for demand response because product temperatures lower than the setpoint (often around 0⁰F) are typically acceptable. Refrigerated beverages also have good potential, because of a relatively large range of acceptable product temperatures. Perishables in coolers (>32⁰F) such as fruits, dairy, and meat are typically not good candidates for demand response because these products have a very narrow range of acceptable storage temperatures.
Additional end-uses besides refrigeration may also have demand response opportunity. These include battery charging for forklifts and pallet drives, HVAC and lighting in office spaces, and process loads.
MECHANISMS
Refrigeration loads can be reduced by reducing compressor, condenser fan, and/or evaporator fan usage. Depending on the controls of the system, reductions may be automated at a very high level, eliciting an orchestrated response of the various refrigeration system components, or may be applied to one or more components of the system. Components on variable speed drives may be run at lower speeds, otherwise they may be cycled on and off.
Infiltration through loading dock doors is often large driver of refrigeration loads. Infiltration can be temporarily reduced by employee behavioral changes such as keeping doors shut when not in use and minimizing door open times. Additional savings are available year-round from automated doors and other technologies.
Additional refrigeration system load reduction can be achieved by rescheduling electric reheat on freezer evaporator coils.
STOCK CHARACTERISTICS
Prior work for the Council identified approximately 160 refrigerated warehouses in the Northwest used for cold storage (~75%) and for distribution centers (~25%). There are additional sites for food and ice processing, but they were not the focus of the Council study or this demand response analysis. The work included estimates of annual energy consumption as a function of site type and size; these estimates are based on consumption data available for select sites and publicly available information regarding the size of all sites. The refrigerated warehouses identified are located primarily along the I-5 corridor and in the tri-cities area. These facilities range in size from 10,000s to 100,000s of square feet (median: 156,000), and have estimated annual loads up to about 1.5 aMW (median 0.3). A report summarizing this work is available here.
IMPLEMENTATION COST
Implementation costs range from negligible (for manual adjustments) to $10,000s for automated systems. For example, a Bonneville automated demand response pilot study from 2010-2012 found hardware costs of approximately $10,000 and installation and programming of the hardware to cost approximately $20,000. A case study conducted by the Lawrence Berkeley National Laboratory observed cost of approximately $20,000 for equipment and $35,000 for installation, programming, and monitoring.