The operations reliability of a datacentre is dependent on numerous of factors that all have their effect on each other. Besides the reliability, the energy consumption has to comply with standards and regulation, not to speak of the possible reduction on the running costs. By simulations the effect of either a single or a series of measures can be evaluated in terms of performance, liability, running costs reduction and payback time.
External influences and equipment performance
Datacentre cooling is essential for its operation reliability and should always be available during all different climate conditions. Cooling generation equipment is mostly installed on the roof of the facility in close proximity of heat generating equipment as generator exhaust systems, air handling units and more. Recent studies showed that heat dissipation from the different sources on the roof as well as solar irradiation does influence the efficiency. Read more about this study from our datacentre presentation sheet.
Depending on the usage, the hardware in a datacentre room produces a lot of heat which varies per time and/or location. The cooling system must be able to adapt to the local cooling demand, if not hot spots can occur which can cause damage. The performance of the hardware is only guaranteed within certain temperature limits. Recently we investigated the possibilities of a fanless dataroom cooling concept in collaboration with Tebodin and Rittal. Read more in Datacentreworks.
Equipment performance assessment and optimization
Besides the global performance of the cooling system on the roof and the ventilation distribution in the dataroom itself there are numerous other system components on whose performance the reliability depends. For example a buffer tank in the cooling system should have a clear stratification of hot and cold water. This ensures the supply water temperature for CRAC units over the required time in case of power failures.