Modeling and Control for Cooling Management of Data Centers With Hot Aisle Containment

Abstract

In traditional raised-floor data center design with hot aisle and cold aisle separation, the cooling efficiency suffers from recirculation resulting from the mixing of cool air from the Computer Room Air Conditioning (CRAC) units and the hot exhaust air exiting from the back of the server racks. To minimize recirculation and hence increase cooling efficiency, hot aisle containment has been employed in an increasing number of data centers. Based on the underlying heat transfer principles, we present in this paper a dynamic model for cooling management in both open and contained environment, and propose decentralized model predictive controllers (MPC) for control of the CRAC units. One approach to partition a data center into overlapping CRAC zones of influence is discussed. Within each zone, the CRAC unit blower speed and supply air temperature are adjusted by a MPC controller to regulate the rack inlet temperatures, while minimizing the cooling power consumption. The proposed decentralized cooling control approach is validated in a production data center with hot aisles contained by plastic strips. Experimental results demonstrate both its stability and ability to reject various disturbances.

Energy efficiency and renewable energy integration in data centres. Strategies and modelling review

“…Raising the set point temperature in the data centre as well as hot and cold aisle containments were also found to have significant effect on the power consumption. Zhou et al [100] developed a numerical model which allowed capturing the effects of the fan speed of CRAC units and supply air temperatures on rack inlet temperatures. The developed model permits analysing different aspects of the cooling system design such as thermal zone mapping, CRAC units load balancing and hot spot detection.…”

Optimization and control of cooling microgrids for data centers

“…For air-cooled data centers with raised floors, general dynamic models have been developed in the authors' previous work [4], [6]. The dynamics of a single rack inlet air temperature T at discrete time step k and k + 1, for example, is described by the following equation [4]: (1) in which SAT i and V F D i are the supply air temperature and blower speed of the i th CRAC unit, and U j is the normalized opening of the j th vent tile.…”

“…1, various controller design techniques such as model predictive control (MPC) can be used to coordinate the various cooling actuations and maintain the data center at the target thermal status while minimizing the cost function. The interested reader can refer to [4], [6] for detailed MPC controller design for data center cooling control.…”

“…DSC lacked an explicit cooling optimization capability, which has been addressed with physics-based dynamic models for data center cooling [4], [5]. The computationally efficient data center cooling model developed in [4], for example, leads to optimal data center cooling power control using both centralized [4] and decentralized [6] control system architectures for minimized cooling power. Model-based data center cooling analysis [7] can also be performed for improved thermal zone mapping and local hot spot detection.…”

“…The strength of this influence, however, can vary tremendously depending on the relative location between a specific CRAC unit and the rack inlet. A thermal zone can thus be defined for each CRAC unit, denoting the region where it has the most significant influence over the rack inlet air temperatures [3], [6], [7]. The cool supply air in the under-floor plenum is delivered to the rack inlets through the perforated vent tiles in the cold aisles due to the pressure difference beneath and above the floor.…”