No abstract

Executive Summary KGS Buildings LLC (KGS) and Pacific Northwest National Laboratory (PNNL) have developed a simplified control algorithm and prototype low-lift chiller controller suitable for model-predictive control in a demonstration project of low-lift cooling. Low-lift cooling is a highly efficient cooling strategy conceived to enable low or net-zero energy buildings (Armstrong et al. 2009a, b). A low-lift cooling system consists of a high efficiency low-lift chiller, radiant cooling, thermal storage, and model-predictive control to pre-cool thermal storage overnight on an optimal cooling rate trajectory. We call the properly integrated and controlled combination of these elements a low-lift cooling system (LLCS).An effective design approach for low-lift cooling is to use building integral thermal mass for thermal storage. This can be accomplished with thermo-active building systems (TABS), where pipe is embedded in concrete to allow for active charging of building thermal mass. This approach can be applied easily to new construction, or to existing buildings undergoing a deep retrofit, where a thin "topping" slab can be applied to existing infrastructure. . A low lift chiller is required to achieve high precooling efficiencies. A low-lift chiller is a chiller than can operate efficiently over a wide range of compressor speed and condensing-evaporating temperature difference. This ability results in very high part-load efficiencies because at low compressor speed not only are flow losses reduced but condenser and evaporator approach temperatures are also improved, i.e. internal temperature lift approaches external temperature lift.This work addresses two barriers to achieving commercially viable low-lift cooling systems: 1) the lack of chillers that operate efficiently at part-load, or low pressure ratios, and which can be properly controlled in the precooling mode, and 2) the need for a multi-zone model-predictive control algorithm that can be implemented through a commercial building automation system (BAS). The first barrier can be overcome by modifying the controls interface in an appropriate variable refrigerant flow (VRF) chiller to receive high-level logic commands from a supervisory control algorithm. These modifications will allow operation at a specified load schedule, where compressor speed and condenser (or cooling tower) air flow rates are optimized to minimize energy consumption over 24 hours and at each temperature and load condition. The second barrier arises because BASs do not allow complex optimization or model identification methods to be incorporated into building controls. In the foreseeable future, a simplified control algorithm that does not require complicated optimization solvers is needed to demonstrate low-lift cooling systems. This research addresses these barriers through the following four tasks:

No abstract